Received: 11 April 2018 Accepted: 31 May 2018 DOI: 10.1111/pedi.12701

ISPAD CLINICAL PRACTICE CONSENSUS GUIDELINES

ISPAD Clinical Practice Consensus Guidelines 2018: Diabetic ketoacidosis and the hyperglycemic hyperosmolar state

Joseph I. Wolfsdorf1 | Nicole Glaser2 | Michael Agus1,3 | Maria Fritsch4 | Ragnar Hanas5 | Arleta Rewers6 | Mark A. Sperling7 | Ethel Codner8

1Division of Endocrinology, Boston Children's Hospital, Boston, Massachusetts 2Department of Pediatrics, Section of Endocrinology, University of California, Davis School of Medicine, Sacramento, California 3Division of Critical Care Medicine, Boston Children's Hospital, Boston, Massachusetts 4Department of Pediatric and Adolescent Medicine, Medical University of Vienna, Vienna, Austria 5Department of Pediatrics, NU Hospital Group, Uddevalla and Sahlgrenska Academy, Gothenburg University, Uddevalla, Sweden 6Department of Pediatrics, School of Medicine, University of Colorado, Aurora, Colorado 7Division of Endocrinology, Diabetes and Metabolism, Department of Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York 8Institute of Maternal and Child Research, School of Medicine, University of Chile, Santiago, Chile Correspondence Joseph I. Wolfsdorf, Division of Endocrinology, Boston Children's Hospital, 300 Longwood Avenue, Boston, MA. Email: [email protected]

1 | SUMMARY OF WHAT IS Risk factors for DKA in newly diagnosed patients include younger NEW/DIFFERENT age, delayed diagnosis, lower socioeconomic status, and residence in a country with a low prevalence of type 1 diabetes mellitus (T1DM). Recommendations concerning fluid management have been modified Risk factors for DKA in patients with known diabetes include to reflect recent findings from a randomized controlled clinical trial omission of insulin for various reasons, limited access to medical ser- showing no difference in cerebral injury in patients rehydrated at dif- vices, and unrecognized interruption of insulin delivery in patients ferent rates with either 0.45% or 0.9% saline. using an insulin pump. The following recommendations are based on currently avail- able evidence and are intended to be a general guide to DKA man- agement. Because there is considerable individual variability in 2 | EXECUTIVE SUMMARY presentation of DKA (ranging from mild with only minimal dehydra- tion to severe with profound dehydration), some patients may The biochemical criteria for the diagnosis of diabetic ketoacidosis require specific treatment that, in the judgment of the treating phy- (DKA) are: sician, may be within or, occasionally, outside the range of options presented here. Clinical judgment should always be used to deter- • Hyperglycemia (blood glucose >11 mmol/L [≈200 mg/dL]) mine optimal treatment for the individual patient, and timely • Venous pH <7.3 or serum bicarbonate <15 mmol/L adjustments to treatment (electrolyte composition and rate of infu- • Ketonemia (blood ß-hydroxybuyrate ≥3 mmol/L) or moderate or sion of rehydration fluids, insulin dose) should be based on ongo- large ketonuria. ing, careful clinical and biochemical monitoring of the patient's response. The clinical signs of DKA include: Dehydration, tachycardia, Emergency assessment should follow the general guidelines tachypnea, deep sighing respiration, breath smells of acetone, nausea for Pediatric Advanced Life Support (PALS) and includes: Imme- and/or vomiting, abdominal pain, blurry vision, confusion, drowsiness, diate measurement of blood glucose, blood or urine ketones, progressive decrease in level of consciousness and, eventually, loss of serum electrolytes, blood gases and complete blood count; consciousness (coma). assessment of severity of dehydration, and level of

© 2018 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Pediatric Diabetes October 2018; 19 (Suppl. 27): 155–177. wileyonlinelibrary.com/journal/pedi 155 156 WOLFSDORF ET AL. consciousness (E). A second peripheral intravenous (IV) catheter • Plasma glucose concentration >33.3 mmol/L (600 mg/dL) should be inserted (E). • Venous pH >7.25; arterial pH >7.30 Management should be conducted in a center experienced in • Serum bicarbonate >15 mmol/L the treatment of DKA in children and adolescents and where vital • Small ketonuria, absent to mild ketonemia signs, neurological status and laboratory results can be monitored • Effective serum osmolality >320 mOsm/kg frequently (E). Where geographic constraints require that manage- • Altered consciousness (eg, obtundation, combativeness) or sei- ment be initiated in a center with less experience and with fewer zures (in approximately 50%) resources, there should be arrangements in place for telephone or videoconference support from a physician with expertise in In HHS, the goals of initial fluid therapy are to expand the intra- DKA (E). and extravascular volume, restore normal renal perfusion and promote Meticulous monitoring of the clinical and biochemical response a gradual decline in corrected serum sodium concentration and serum to treatment is necessary so that timely adjustments in treatment osmolality. can be made when indicated by the patient's clinical or laboratory In HHS, begin insulin administration at a dose of 0.025 to data (E). 0.05 U/kg/h once plasma glucose is decreasing less than 3 mmol/L Goals of therapy are to correct dehydration, correct acidosis and (50 mg/dL) per hour with fluid alone (C). DKA results from deficiency of circulating insulin and increased reverse ketosis, gradually restore hyperosmolality and blood glucose levels of the counterregulatory hormones: catecholamines, glucagon, concentration to near normal, monitor for complications of DKA and cortisol, and growth hormone.1,2 Severe insulin deficiency occurs in its treatment, and identify and treat any precipitating event. previously undiagnosed T1DM and when treated patients deliber- Fluid replacement should begin before starting insulin therapy. ately or inadvertently do not take insulin, especially the long-acting Expand volume using crystalloids, as required, to restore peripheral component of a basal-bolus regimen, or markedly reduce the doses circulation (E). Calculate the subsequent rate of fluid administration, of insulin, for example, in association with an intercurrent illness including the provision of maintenance fluid requirements, aiming to such as gastroenteritis. Patients who use an insulin pump can rap- replace the estimated fluid deficit over 24 to 48 hours (A). idly develop DKA when insulin delivery fails for any reason.3 Rela- Insulin therapy: begin with 0.05 to 0.1 U/kg/h at least 1 hour tive insulin deficiency occurs when the concentrations of AFTER starting fluid replacement therapy (B). counterregulatory hormones markedly increase in response to stress Potassium: If the patient is hyperkalemic, defer potassium replace- in conditions such as sepsis, trauma, or febrile illness, which over- ment therapy until urine output is documented. Otherwise, begin with whelm homeostatic mechanisms and lead to metabolic decompensa- 40 mmol potassium/L (or 20 mmol potassium/L if the patient is tion despite the patient taking the usual recommended dose of receiving fluid at a rate ≥10 mL/kg/h) (E). insulin. Bicarbonate administration is not recommended except for The combination of absolute or relative insulin deficiency and treatment of life-threatening hyperkalemia or unusually severe high counterregulatory hormone concentrations causes an acceler- acidosis (vpH <6.9) with evidence of compromised cardiac ated catabolic state with increased glucose production by the liver contractility (C). and kidney (via glycogenolysis and gluconeogenesis) and impaired Warning signs and symptoms of cerebral edema include: Onset peripheral glucose utilization, which result in hyperglycemia and of headache after beginning treatment or progressively worsening or hyperosmolality. Insulin deficiency and high counterregulatory hor- severe headache, slowing of heart rate not related to sleep or mone concentrations also increase lipolysis and ketogenesis and improved intravascular volume, change in neurological status (restless- cause ketonemia and metabolic acidosis. Hyperglycemia exceeding ness, irritability, increased drowsiness, confusion, incontinence), spe- the usual renal threshold of approximately 10 mmol/L (180 mg/dL) cific neurological signs (eg, cranial nerve palsies), rising blood pressure, together with hyperketonemia cause osmotic diuresis, dehydration, and decreased oxygen saturation (C). and obligatory loss of electrolytes, often aggravated by vomiting In patients with multiple risk factors for cerebral edema (ele- associated with severe ketosis. These changes stimulate further vated serum urea nitrogen concentration, severe acidosis, severe stress hormone production, which induces more severe insulin resis- hypocapnia), have mannitol or hypertonic saline at the bedside tance and worsening hyperglycemia and hyperketonemia. Lactic aci- and the dose calculated (E). If neurologic status deteriorates dosis from hypoperfusion or sepsis may contribute to the acidosis acutely, hyperosmolar fluid therapy should be given immedi- (Figure 1).4 ately (C). If this cycle is not interrupted by exogenous insulin together with Prevention: Management of an episode of DKA is not complete fluid and electrolyte therapy, fatal dehydration and metabolic acidosis until an attempt has been made to identify and treat the cause. will ensue. DKA without a preceding febrile illness or gastroenteritis in a DKA is characterized by severe depletion of water and electro- patient with known diabetes is almost always the result of psychoso- lytes from both the intra- and extracellular fluid (ECF) compart- cial problems and failure to appropriately administer insulin. ments5; the typical range of losses is shown in Table 1. Despite In new onset diabetes, DKA is frequently the consequence of a their dehydration, patients generally continue to maintain normal delay in diagnosis (E). blood pressure or even have high blood pressure,6 possibly due to The criteria for hyperglycemic hyperosmolar state (HHS) include: elevated plasma catecholamine concentrations, increased release of WOLFSDORF ET AL. 157

Pathophysiology of Diabetic Ketoacidosis

Absolute insulin deficiency or Stress, infection or insufficient insulin

Counterregulatory Hormones ↑ Glucagon ↑ Cortisol ↑ Catecholamines ↑ Growth Hormone

↑ Lipolysis ↓Glucose utilization ↑ Proteolysis ↑ Glycogenolysis ↓ Protein synthesis

↑ Gluconeogenic substrates

↑ FFA to liver ↑ Gluconeogenesis

↑ Ketogenesis Hyperglycemia

↓ Alkali reserve Glucosuria (osmotic diuresis)

Acidosis Loss of water and electrolytes Decreased fluid intake ↑ Lactate Dehydration Hyperosmolarity

Impaired renal function

FIGURE 1 Pathophysiology of diabetic ketoacidosis. Copyright© 2006 American Diabetes Association. From diabetes care, Vol. 29, 2006:1150-1159. Reprinted with permission of The American Diabetes Association antidiuretic hormone (ADH) in response to hyperosmolality (which TABLE 1 Losses of fluid and electrolytes in diabetic ketoacidosis and increases blood pressure via V2 receptors), increased osmotic pres- maintenance requirements in normal children sure from marked hyperglycemia, or other factors.6 Considerable Average (range) urine output persists because of glucosuria until extreme volume losses per kg 24-hour maintenance requirements * depletion leads to a critical decrease in renal blood flow and glo- Water 70 mL (30-100) ≤10 kg 100 mL/kg/24 h merular filtration. At presentation, the specific deficits in an individ- 11-20 kg 1000 mL + 50 mL/kg/24 h for each kg from 11 to 20 ual patient vary depending upon the duration and severity of >20 kg 1500 mL + 20 mL/kg/24 h illness, the extent to which the patient was able to maintain intake for each kg >20 of fluid and electrolytes, and the content of food and fluids con- † Sodium 6 mmol (5-13) 2-4 mmol sumed before coming to medical attention. Consumption of fluids Potassium 5 mmol (3-6) 2-3 mmol with a high-carbohydrate content (juices or sugar-containing soft Chloride 4 mmol (3-9) 2-3 mmol drinks) may exacerbate the hyperglycemia.7 Conversely, though Phosphate 0.5-2.5 mmol 1-2 mmol uncommonly, modest hyperglycemia in the setting of severe acidosis may be an indication that the patient has maintained increased water intake and may be only modestly hypovolemic. Rapid empty- ing of stomach contents containing an abundant quantity of sugar, Data are from measurements in only a few children and which occurs as gastroparesis is relieved with therapy, accounts for adolescents.9–13 In any individual patient, actual losses may be less or the rise in plasma glucose concentration observed in some patients more than the ranges shown in Table 1. after onset of therapy despite ongoing large loss of glucose in the Three methods for determining maintenance water requirements urine.8 in children are commonly used: *the Holliday-Segar formula14 (shown in Table 1), a simplified Holliday-Segar formula (see below), and a for- mula based on body surface area for children who weigh more than Clinical manifestations of diabetic ketoacidosis 10 kg (1500 mL/m2/24 h).15 † (shown in Table 1) Maintenance electrolyte requirements in • Dehydration children are per 100 mL of maintenance IV fluid.15,16 • Tachypnea; deep, sighing (Kussmaul) respiration Simplified method based on Holliday-Segar: <10 kg 4 mL/kg/h; • Nausea, vomiting, and abdominal pain that may mimic 11-20 kg 40 + 2 mL/kg/h for each kg between 11 and 20; >20 kg an acute abdominal condition 60 + 1 mL/kg/h for each kg >20. • Confusion, drowsiness, progressive obtundation, and To avoid excessive fluid administration in obese patients, fluid cal- loss of consciousness culations should be based on an approximation of ideal body weight for height. 158 WOLFSDORF ET AL.

TABLE 2 Fluid maintenance and replacement volumes based on • Hyperglycemia (blood glucose >11 mmol/L [200 mg/dL]) body weight and an assumption of 10% dehydration • Venous pH <7.3 or serum bicarbonate <15 mmol/L DKA: give maintenance • Ketonemia* or ketonuria. +5% of body weight/24 h Body weight Maintenance (kg) (mL/24 h) mL/24 h mL/h *Although not universally available, blood beta-hydroxybutyrate 4 325 530 22 (BOHB) concentration should be measured whenever possible; a level 5 405 650 27 ≥3 mmol/L is indicative of DKA.19 6 485 790 33 Urine ketones are typically ≥2+ “moderate or large” positive. Par- 7 570 920 38 tially treated children and children who have consumed little or no 8 640 1040 43 carbohydrate may, rarely, have only modestly elevated blood glucose 9 710 1160 48 concentrations, referred to as “euglycemic ketoacidosis”.20,21 This can 10 780 1280 53 be caused by starvation (anorexia or religious fasting),22 a low carbo- 11 840 1390 58 hydrate high fat diet, or the off-label use of SGLT2-inhibitors.23–25 12 890 1490 62 Serum bicarbonate concentration alone can substitute for vpH to 13 940 1590 66 diagnose DKA and classify severity in children with new onset diabe- 14 990 1690 70 tes mellitus and is suggested as an alternative to reliance on vpH in 15 1030 1780 74 settings where access to vpH measurement is limited.26 16 1070 1870 78 The frequency of type 2 diabetes mellitus (T2DM) in the pediatric age 17 1120 1970 82 range is increasing.27,28 The worldwide incidence and prevalence of type 18 1150 2050 85 2 diabetes in children and adolescents vary substantially among countries, 19 1190 2140 89 age categories, and ethnic groups, which can be explained by variations in 20 1230 2230 93 population characteristics and methodological dissimilarities between stud- 22 1300 2400 100 ies.29 DKA at diagnosis is more common in younger children, minority 24 1360 2560 107 race, and male gender.30 At some centers in the United States, type 2 dia- 26 1430 2730 114 betes now accounts for approximately one-half of newly diagnosed diabe- 28 1490 2890 120 tesinchildrenaged10to21years.31 The SEARCH for Diabetes in Youth 30 1560 3060 128 Study in the United States found that DKA has decreased over time; the 32 1620 3220 134 most recent data show that it occurred in nearly 6% of youth with type 34 1680 3360 140 2diabetes.30,32 Overall, 5% to 25% of patients with type 2 diabetes have 36 1730 3460 144 DKA at the time of diagnosis.33,34 38 1790 3580 149 The severity of DKA is categorized by the degree of acidosis35: 40 1850 3700 154 45 1980 3960 165 • Mild: venous pH <7.3 or serum bicarbonate <15 mmol/L 50 2100 4200 175 • Moderate: pH <7.2, serum bicarbonate <10 mmol/L 55 2210 4420 184 • Severe: pH <7.1, serum bicarbonate <5 mmol/L 60 2320 4640 193 65 2410 4820 201 HHS, formerly referred to as hyperosmolar non-ketotic coma, may 70 2500 5000 208 occur in young patients with T2DM,34,36–38 in type 1 diabetes 75 2590 5180 216 patients,39 and in infants, especially those with 6q24-related transient 80 2690 5380 224 neonatal diabetes mellitus.40 The criteria for HHS include2,41:

After initial resuscitation, and assuming 10% dehydration, the • plasma glucose concentration >33.3 mmol/L (600 mg/dL) total volume of fluid has been calculated to be given over 48 hours. • arterial pH >7.30; venous pH >7.25 The table shows volumes for maintenance and rehydration per • serum bicarbonate >15 mmol/L • * 24 hours and per hour. Fluid given orally (when patient has improved) small ketonuria, absent to small ketonemia • should be subtracted from the volume in the table. Fluid volumes are effective serum osmolality >320 mOsm/kg • calculated based on data from Darrow.17 For body weights >32 kg, obtundation, combativeness, or seizures (in approximately 50%) the volumes have been adjusted so as not to exceed twice the mainte- Overlap between the characteristic features of HHS and DKA nance rate of fluid administration. may occur, and some patients with HHS, especially when there is severe dehydration, have mild or moderate acidosis that is mainly due 3 | DEFINITION OF DIABETIC to hypoperfusion and lactic acidosis. Conversely, some children with KETOACIDOSIS type 1 diabetes may have features of HHS (severe hyperglycemia) especially if high-carbohydrate containing beverages have been used The biochemical criteria for the diagnosis of DKA are18: to quench thirst and replace urinary losses before diagnosis.7 Therapy WOLFSDORF ET AL. 159 must be appropriately modified to address the pathophysiology and • Immediately measure blood glucose and blood BOHB concentrations particular biochemical disturbances of the individual patient (see with bedside meters or with urine test strips that measure only acet- below). See below regarding specific therapy of HHS. oacetic acid if bedside blood ketone measurements are not available. Perform a clinical evaluation to identify a possible infection. 4 | FREQUENCY OF DKA • Measurement of blood BOHB concentration with a point-of-care meter, if available, is useful to confirm ketoacidosis (≥3mmol/Lin children)19 and to monitor the response to treatment.67–73 4.1 | At disease onset • Weigh the patient. If body surface area is used for fluid therapy There is wide geographic variation in the frequency of DKA at onset calculations, measure height or length to determine surface area. of diabetes; rates inversely correlate with the regional incidence of The current weight should be used for calculations and not the type 1 diabetes. Frequencies range from approximately 15% to 70% weight from a previous office visit or hospital record. in Europe and North America.30,42–49 DKA at diagnosis is more com- • Assess severity of dehydration. mon in younger children (especially <2 years of age), including infants • Estimation of the degree of dehydration is imprecise and gen- with both transient and permanent neonatal diabetes (overall fre- erally shows only fair to moderate agreement among – quency 66%), often the consequence of diagnostic error or delayed examiners,74 76 and should be based on a combination of treatment.50–53 It is also more common in ethnic minority groups, and physical signs. The most useful signs for predicting 5% dehy- in children whose families do not have ready access to medical care dration in young children aged 1 month to 5 years are: for social or economic reasons.21,32,46,51,54,55 • prolonged capillary refill time (normal capillary refill is ≤1.5-2 seconds) • “ ” 4.2 | In children with established diabetes abnormal skin turgor ( tenting or inelastic skin) • Other useful signs in assessing the degree of dehydration The risk of DKA in established type 1 diabetes is 1% to 10% per include: dry mucus membranes, sunken eyes, absent tears, 3,56–61 patient per year : weak pulses, cool extremities. More signs of dehydration tend 59 Risk is increased in : to be associated with more severe dehydration.77 • ≥10% dehydration is suggested by the presence of weak or 58 • Children who omit insulin impalpable peripheral pulses, hypotension, oliguria. • Children with poor metabolic control or previous episodes of DKA • Assess level of consciousness (Glasgow coma scale [GCS]—see • Gastroenteritis with persistent vomiting and inability to maintain Table 3).78 hydration • In the unconscious or severely obtunded patient without normal air- • Children with psychiatric disorders, including those with eating way protective reflexes, secure the airway and empty the stomach disorders by continuous nasogastric suction to prevent pulmonary aspiration. • Children with difficult or unstable family circumstances (eg, paren- • Intubation should be avoided if possible; an increase of pCO2 tal abuse) during or following intubation above the level that the patient • Peripubertal and adolescent girls had been maintaining may cause cerebrospinal fluid (CSF) pH 62 • Binge alcohol consumption to decrease and contribute to worsening of cerebral edema.79 • Children with limited access to medical services • If there is a history of recent large consumption of glucose- containing fluids, consider emptying the stomach even in the In the early days of insulin pump therapy, DKA was more common patient who is not obtunded. than in patients using injection therapy (only rapid- or short-acting • When large quantities of fruit juice or sweetened soft drinks insulin is used in pumps; therefore, interruption of insulin delivery for have been ingested, the stomach may contain a large volume 3,63 any reason rapidly leads to insulin deficiency). However, a recent of water with little sodium. Spontaneous gastric emptying early matched comparison of patients using insulin pump therapy with mul- in the course of therapy leads to absorption of glucose and tiple daily injections showed that DKA occurred less frequently (3.64 electrolyte-free water from the intestinal tract.8,80 vs 4.26 per 100 patient-years) in patients using pump therapy.64 • Give oxygen to patients with circulatory impairment or shock. In recurrent DKA, insulin omission or failure to follow sick day or • A cardiac monitor should be used for continuous electrocardio- pump failure management guidelines accounts for almost all episodes. graphic monitoring to assess T-waves for evidence of hyper- or hypokalemia.81,82 • 5 | MANAGEMENT OF DKA A second peripheral IV catheter should be placed for convenient and painless repetitive blood sampling. An arterial catheter may,

Figure 2 shows an algorithm for the management of DKA. rarely, be necessary in some critically ill patients managed in an intensive care unit. • Unless absolutely necessary, avoid placing a central venous 5.1 | Emergency assessment catheter because of the high risk of thrombosis, especially in Acute management should follow the general guidelines for the very young child. If a central catheter has been inserted, PALS,65,66 with particular attention to the following: the catheter should be removed as soon as the patient's clinical 160 WOLFSDORF ET AL.

Biochemical features & Immediate assessment Clinical Signs investigations Dehydration Ketones in urine Deep sighing ↑ blood glucose Clinical History respiration (Kussmaul) Polyuria, polydipsia, Smell of ketones Acidemia (vpH <7.3, HCO3 <15 nocturia, enuresis Lethargy/drowsiness mmol/L) Weight loss Urea, electrolytes Nausea, vomiting, Other investigations as indicated abdominal pain Weakness, fatigue Confusion, ↓ level of Diagnosis confirmed consciousness Diabetic Ketoacidosis Contact Senior Staff

Shock (reduced peripheral pulses) Dehydration >5% Minimal dehydration Reduced conscious level/coma Not in shock Tolerating oral fluid Acidotic (hyperventilation) Vomiting

Resuscitation IV Therapy Therapy Airway ± NG tube Saline 0.9% 10 mL/kg over 1 h; may repeat Start with SC insulin Breathing (100% oxygen) Calculate fluid requirements Continue oral hydration Circulation (0.9% saline Correct fluid deficit over 24-48 hours 10-20 ml/kg over 1-2 h. ECG for abnormal T-waves & repeat until circulation is Add KCl 40 mmol per liter fluid restored See CE Management No improvement

Continuous insulin infusion at 0.05 - 0.1 unit/kg/hour starting 1 hour after fluids initiated

Critical Observations Hourly blood glucose Hourly fluid input & output Neurological status at least hourly Electrolytes 2 hourly after starting IV fluid therapy Monitor ECG for T-wave changes

Acidosis not improving Blood glucose ≤17 mmol/L (300 mg/dL) Neurological deterioration or WARNING SIGNS: Blood glucose falls >5 mmol/L/hour (90 mg/dL) severe or progressive headache, slowing heart rate, irritability, confusion, Re-evaluate decreased consciousness, IV Therapy incontinence, specific IV fluid calculations Change to 0.45% or 0.9% saline; add glucose neurological signs Insulin delivery system & dose to fluids (5% - 12.5%) to prevent hypoglycemia Need for additional resuscitation Adjust sodium infusion to promote an Consider sepsis increase in measured serum sodium Exclude hypoglycemia Is it cerebral edema?

Improved CE Management Clinically well,ketoacidosis resolved, tolerating oral fluids Give mannitol 0.5-1g/kg or 3% hypertonic saline Transition to SC Insulin Adjust IV fluids to maintain Start SC insulin then stop IV insulin after an normal BP, but avoid over- appropriate interval hydration Call senior staff Move to ICU Consider cranial imaging only after patient stabilised

FIGURE 2 Algorithm for the management of diabetic ketoacidosis. Adapted from Pinhas-Hamiel and Sperling.271 NG, nasogastric; SC, subcutaneous WOLFSDORF ET AL. 161

TABLE 3 Glasgow coma scale or score (GCS)

Best verbal response (non-verbal Best eye response Best verbal response children) Best motor response 1. No eye opening 1. No verbal response 1. No response 1. No motor response 2. Eyes open to pain 2. No words, only incomprehensible 2. Inconsolable, irritable, restless, cries 2. Extension to pain 3. Eyes open to verbal sounds; moaning 3. Inconsistently consolable and moans; (decerebrate posture) command 3. Words, but incoherenta makes vocal sounds 3. Flexion to pain (decorticate 4. Eyes open 4. Confused, disoriented conversationb 4. Consolable when crying and interacts posture) spontaneously 5. Oriented, normal conversation inappropriately 4. Withdrawal from pain 5. Smiles, oriented to sound, follows 5. Localizes pain objects and interacts 6. Obeys commands

The GCS consists of three parameters and is scored between 3 and 15; 3 being the worst and 15 the best.78 One of the components of the GCS is the best verbal response, which cannot be assessed in non-verbal young children. A modification of the GCS was created for children too young to talk. a Inappropriate words, random or exclamatory articulated speech, but no sustained conversational exchange. b Attention can be held; patient responds to questions coherently, but there is some disorientation and confusion.

status permits.83–85 Mechanical and pharmacologic prophylaxis • Experienced nursing staff trained in monitoring and management (low molecular weight heparin) should be considered especially of DKA in children and adolescents in children >12 years. • Written guidelines or, if unavailable, access to online guidelines • Insulin should preferably not be given through a central line for DKA management in children unless it is the only available option because its infusion may • Access to a laboratory that can provide frequent and timely mea- be interrupted when other fluids are given through the surements of biochemical variables same line. • Give antibiotics to febrile patients after obtaining appropriate Whenever possible, a specialist/consultant pediatrician with train- cultures of body fluids. ing and expertise in the management of DKA should direct inpatient • Bladder catheterization usually is not necessary, but if the child is management. Where geographic constraints require that management unconscious or unable to void on demand (eg, infants and very ill be initiated in a center with less experience and with fewer resources, young children) the bladder should be catheterized. there should be arrangements in place for telephone or videoconfer- • Obtain a blood sample for laboratory measurement of: ence support from a physician with expertise in DKA. • serum or plasma glucose Children with severe DKA (long duration of symptoms, compro- • electrolytes (including serum bicarbonate) mised circulation, or depressed level of consciousness) or those who • blood urea nitrogen, creatinine are at increased risk for cerebral edema (eg, <5 years of age, severe • serum osmolality acidosis, low pCO2, high blood urea nitrogen) should be considered • † venous pH, pCO2 for immediate treatment in an intensive care unit (pediatric if avail- • hemoglobin, hematocrit and complete blood count. Note that able) or in a unit that has equivalent resources and supervision, such an increased white blood cell count in response to stress is as a children's ward specializing in diabetes care.18,87 Transport teams 86 characteristic of DKA and is not indicative of infection. should be knowledgeable about DKA management (or have access to • albumin, calcium, phosphate, magnesium concentrations a medical control physician who is knowledgeable) and should have (if possible) rescue medications available during the transport, including high con- • Perform a urinalysis for ketones if blood or serum ketones have centration dextrose IV solutions and mannitol or 3% hypertonic saline. not been measured. In a child with established diabetes, whose parents have been • Obtain appropriate specimens for culture (blood, urine, throat), trained in sick day management, hyperglycemia and ketosis without only if there is evidence of infection (eg, fever). vomiting or severe dehydration can be managed at home or in an out- • If laboratory measurement of serum potassium is delayed, per- patient health care facility (eg, emergency ward), provided an experi- form an electrocardiogram (ECG) for baseline evaluation of potas- enced diabetes team supervises the care.35,88,89 sium status.81,82 • Although not essential for management of DKA per se, HbA1c may be useful in the evaluation and management of specific patients as 7 | CLINICAL AND BIOCHEMICAL it provides information about the duration of hyperglycemia. MONITORING

Successful management of DKA and HHS requires meticulous moni- toring and recording of the patient's clinical and biochemical response 6 | WHERESHOULDTHECHILDWITHDKA to treatment so that timely adjustments in treatment can be made BE MANAGED? when indicated by the patient's clinical or laboratory data. There should be documentation on a flow chart of hour-by-hour After initial life support, the child should receive care in a unit clinical observations, IV and oral medications, fluids, and laboratory that has: results. Monitoring should include the following: 162 WOLFSDORF ET AL.

• Hourly (or more frequently as indicated) vital signs (heart rate, respiratory rate, blood pressure) Goals of therapy • Hourly (or more frequently as indicated) neurological observa- tions (Glasgow coma score; Table 3) for warning signs and symp- • Correct acidosis and reverse ketosis toms of cerebral edema (see below) • Correct dehydration • onset of headache after starting DKA treatment or worsening • Restore blood glucose to near normal of headache already present before commencing treatment • Monitor for complications of DKA and its treatment • inappropriate slowing of heart rate • Identify and treat any precipitating event • recurrence of vomiting • change in neurological status (restlessness, irritability, increased drowsiness, confusion, incontinence) or specific neu- rologic signs (eg, cranial nerve palsies, abnormal pupillary 7.1 | Fluids and salt responses) • rising blood pressure Patients with DKA have a deficit in ECF volume that usually is in the • decreased oxygen saturation range 5% to 10% of body weight.9,10 Shock with hemodynamic com- • rapidly increasing serum sodium concentration suggesting loss promise is rare in pediatric DKA. Clinical estimates of the volume defi- of urinary free water as a manifestation of diabetes insipidus cit are subjective and inaccurate74–76; therefore, in moderate DKA (from interruption of blood flow to the pituitary gland due to assume 5% to 7% and in severe DKA 7% to 10% dehydration. The cerebral herniation) effective osmolality (formula above) is frequently in the range of • Amount of administered insulin 300 to 350 mmol/kg. Increased serum urea nitrogen and hematocrit • Hourly (or more frequently as indicated) accurate fluid input or hemoglobin concentration or, alternatively, plasma albumin or total (including all oral fluid) and output. protein concentration if anemia is suspected94 are useful markers of • Capillary blood glucose concentration should be measured hourly the degree of ECF contraction,89,95,96 and should be determined fre- (but must be cross-checked against laboratory venous glucose as quently during fluid resuscitation and deficit replacement.97 The capillary methods may be inaccurate in the presence of poor serum sodium concentration is an unreliable measure of the degree of peripheral circulation and acidosis, and is limited in measuring ECF contraction for two reasons: (1) glucose, largely restricted to the extremely high levels). extracellular space, causes osmotic movement of water into the extra- • Laboratory tests: serum electrolytes, glucose, blood urea nitrogen, cellular space thereby causing dilutional hyponatremia,98,99 and (2) the calcium, magnesium, phosphate, hematocrit, and blood gases low sodium content of the elevated lipid fraction of the serum in should be repeated 2 to 4 hourly, or more frequently, as clinically DKA. The latter is not a concern with most modern methods for mea- indicated, in more severe cases. suring sodium concentration. It is useful to calculate the corrected • Blood BOHB concentrations, if available, every 2 to 4 hours.68–72 sodium concentration (using the above formula), to help assess the • Near-patient (also referred to as point-of-care) BOHB mea- magnitude of the deficit of sodium and water.5 The “corrected” surements correlate well with a reference method up to sodium represents the expected serum sodium concentration in the 3 mmol/L, but are not accurate above 5 mmol/L.70,90 absence of hyperglycemia. Changes in the corrected sodium should • Lipids and triglycerides can be grossly elevated causing the blood be monitored throughout the course of therapy. As the plasma glu- sample to show a visible rim of lipids, which can interfere with cose concentration decreases after administering fluid and insulin, the accuracy of laboratory tests.91 measured serum sodium concentration should increase and the • If the laboratory cannot provide timely results, a portable bio- glucose-corrected sodium concentration (formula above) should chemical analyzer that measures serum electrolytes and blood slowly decrease or remain in the normal range. It is important to gases on fingerstick blood samples at the bedside is a useful appreciate that the increase in measured serum sodium concentration adjunct to laboratory-based determinations. Blood glucose and does not indicate a worsening of the hypertonic state. A failure of blood or urine ketone concentrations can be measured with a measured serum sodium levels to rise or a further decline in serum bedside meter while awaiting results from the laboratory. • Measure body weight each morning. sodium levels with therapy is thought to be a potentially ominous sign 100–102 • Calculations: of impending cerebral edema. A rapid and ongoing rise in serum sodium concentration may also indicate possible cerebral edema as a result of loss of free water in the urine from diabetes • Anion gap = Na − (Cl + HCO3): normal is 12 Æ 2 mmol/L • In DKA the anion gap is typically 20 to 30 mmol/L; an anion insipidus. gap >35 mmol/L suggests concomitant lactic acidosis (E) The objectives of fluid and electrolyte replacement therapy • Corrected sodium = measured Na + 2([plasma glucose − are to: 5.6]/5.6) mmol/L or measured Na + 2([plasma glucose − 100]/100) mg/dL‡ • Restore circulating volume • Effective osmolality (mOsm/kg) = 2 × (plasma Na) + plasma • Replace sodium and the extracellular and intracellular water glucose mmol/L93; normal range is 275 to 295 mOsm/kg deficits WOLFSDORF ET AL. 163

• Improve glomerular filtration and enhance clearance of glucose • Use crystalloid not colloid. There are no data to support the use and ketones from the blood of colloid in preference to crystalloid in the treatment of DKA.

7.2 | Principles of water and salt replacement 7.2.2 | Deficit replacement fluids Despite much effort to identify the cause of cerebral edema, its path- Subsequent fluid management (deficit replacement) can be accom- ogenesis is incompletely understood and controversy continues con- plished with 0.45% to 0.9% saline or a balanced salt solution (Ringer's cerning the association between the rate of fluid or sodium – lactate, Hartmann's solution or Plasmalyte).95,100,108 114 administration used in the treatment of DKA and the development of 103–105 cerebral edema. No treatment strategy can be definitively • Fluid therapy should begin with deficit replacement plus main- recommended as being superior to another based on current evi- tenance fluid requirements. 104 dence. A recently completed prospective randomized clinical trial • All children will experience a decrease in vascular volume (the PECARN FLUID Study) compared acute and long-term neurologi- when plasma glucose concentrations fall during treatment; cal outcomes in 1389 episodes of DKA in 1255 children treated with therefore, it is essential to ensure that they receive suffi- slower vs more rapid fluid administration using either 0.45% or 0.9% cient fluid and salt to maintain adequate tissue perfusion. 106 saline. The PECARN FLUID Study showed no significant differ- • Deficit replacement should be with a solution that has a tonic- ences in the frequency of either altered mental status or clinical diag- ity in the range 0.45% to 0.9% saline, with added potassium noses of cerebral edema in any of the treatment arms, and long-term chloride, potassium phosphate or potassium acetate (see neurocognitive outcomes were similar in all groups. Point estimates below under potassium replacement).95,100,108,112,113,115–117 suggested lower frequencies of altered mental status in children rehy- Decisions regarding use of isotonic vs hypotonic solution for drated more rapidly with 0.45% saline, but these differences did not deficit replacement should depend on clinician judgment based 106 reach statistical significance. The results of this study suggest that on the patient's hydration status, serum sodium concentration an assumed fluid deficit between 5% and 10% of body weight should and osmolality. be replaced over 24 to 48 hours along with maintenance fluids, using • In addition to providing the usual daily maintenance fluid fluids with a sodium content between 0.45% and 0.9% saline. The risk requirement, replace the estimated fluid deficit at an even rate of cerebral injury does not appear to be associated with differences in over 24 to 48 hours.18,95,118 Except for severely ill individuals, fluid protocols within these ranges. Therefore, clinicians should not oral intake typically begins within 24 hours.118 Although rehy- unnecessarily restrict fluid administration if clinical signs suggest the dration is generally planned to occur over longer periods, in a need for circulatory volume expansion. study of 635 episodes of DKA the mean time to correction of The principles described below are based on the consensus state- DKA and complete restoration of the circulation was ment from a panel of expert physicians representing the Lawson Wil- 11.6 Æ 6.2 hours. At this point, any remaining deficits were kins Pediatric Endocrine Society (LWPES), the European Society for replenished by oral intake once DKA had resolved and patients Pediatric Endocrinology (ESPE), and the International Society for Pedi- were transitioned to subcutaneous insulin.118 18,107 atric and Adolescent Diabetes (ISPAD) and incorporate the rec- • Satisfactory outcomes have also been reported using an alter- 106 ommendations from the PECARN FLUID Study. native simplified method: After the initial fluid administration of 20 mL/kg of normal saline, 0.675% saline (3/4 normal saline, • Water and salt deficits must be replaced. 115.5 mmol sodium) is infused at 2 to 2.5 times the usual • IV or oral fluids that may have been given in another facility maintenance rate of fluid administration regardless of the before assessment should be factored into calculations of deficit degree of dehydration, and decreased to 1 to 1.5 times the and replacement volumes. maintenance rate after 24 hours, or earlier if acidosis resolved.112 • Clinical assessment of hydration status and calculated effective 7.2.1 | Resuscitation fluids osmolality are valuable guides to fluid and electrolyte therapy. For patients who are volume depleted but not in shock, volume The aim is gradually to reduce serum effective osmolality to expansion (resuscitation) should begin immediately with 0.9% saline normal.97,118,119 There should be a concomitant increase in to restore the peripheral circulation. The volume administered typi- serum sodium concentration as the serum glucose concentra- cally is 10 mL/kg infused over 30 to 60 minutes; however, if tissue tion decreases (sodium should rise by 0.5 mmol/L for each perfusion is poor the initial fluid bolus is given more rapidly (eg, over 1 mmol/L decrease in glucose concentration). 15-30 minutes) and a second fluid bolus may be needed to ensure • Urinary losses should not routinely be added to the calculation adequate tissue perfusion. of replacement fluid, but this may be necessary in rare circumstances. • In the rare patient with DKA in shock, rapidly restore circulatory • The sodium content of the fluid should be increased if mea- volume with isotonic saline in 20 mL/kg boluses infused as sured serum sodium concentration is low and does not rise quickly as possible through a large bore cannula with reassess- appropriately as the plasma glucose concentration ment of circulatory status after each bolus. falls.100,111,119,120 164 WOLFSDORF ET AL.

• The use of large amounts of chloride-rich fluids (combined edema,105,119,143 can precipitate shock by rapidly decreasing with preferential renal excretion of ketones over chloride) osmotic pressure, and can exacerbate hypokalemia may be associated with the rapid development of • The dose of insulin should usually remain at 0.05 to 0.1 unit/kg/h at hyperchloremia121–123 (defined as a ratio of chloride:sodium least until resolution of DKA (pH >7.30, serum bicarbonate [Cl−:Na+] > 0.79124) and hyperchloremic metabolic >15 mmol/L, BOHB <1 mmol/L, or closure of the anion gap), which acidosis.117,122,125–127 invariably takes longer than normalization of blood glucose concen- • The acidifying effect of chloride can mask recognition of trations.144 Monitor venous pH and serum BOHB concentration resolution of ketoacidosis when total base deficit is used to every 2 hours to ensure steady improvement of biochemical param- monitor biochemical improvement.123 eters. If the insulin effect is adequate serum BOHB should decrease • When hyperchloremia develops, a persisting base deficit or by approximately 0.5 mmol/L/h.72 Increase the insulin dose if the low bicarbonate concentration can be erroneously inter- expected rate of biochemical improvement does not occur. preted as being due to ongoing ketosis.128 • If the patient shows marked sensitivity to insulin (eg, some young • To avoid this misinterpretation, measurement of bedside children with DKA, patients with HHS, and some older children BOHB levels will prevent any confusion and can demon- with established diabetes), the insulin dose may be decreased pro- strate that ketoacidosis has resolved. Hyperchloremic aci- vided that metabolic acidosis continues to resolve. For example, if dosis resolves spontaneously. a young child is receiving 0.05 unit/kg/h, it may be necessary to • Although the anion gap is useful to track resolution of keto- reduce the insulin dose to 0.03 unit/kg/h to prevent hypoglyce- sis, it has two limitations in this setting: it is unable to differ- mia despite the addition of IV glucose. entiate a mixed metabolic acidosis (hyperchloremic and • For less severe DKA (pH >7.1-7.2), 0.05 U/kg/h (0.03 U/kg/h for age ketotic), and the degree of hyperchloremic acidosis is not <5 years with mild DKA) is usually sufficient to resolve the acidosis. quantifiable. • Uncontrolled retrospective and observational studies have • Normally the difference between the serum sodium and reported comparable efficacy and safety using 0.05 unit/kg/ chloride concentrations is 30 to 35 mmol/L. To partition h,145,146 and some pediatric centers routinely use this dose for the chloride component of the base deficit, the following treatment of DKA. A recent small RCT in children ≤12 years old formula has been proposed to enable clinicians to track showed that low dose (0.05 unit/kg/h) was comparable to stan- resolution of ketoacidosis at the bedside: Chloride- dard dose (0.1 U/kg/h) with respect to rate of blood glucose induced base deficit = (plasma sodium − plasma chloride decrease and resolution of acidosis; however, there was also no 113 − 32).123 evidence that the higher dose (0.1 U/kg/h) is harmful. • The chloride load can be reduced by not giving potas- • Insulin has an aldosterone-like effect leading to increased urinary 147–151 sium as potassium chloride (use potassium acetate potassium excretion. High doses administered intrave- instead) and by using fluids such Ringer's lactate or Plas- nously for a prolonged period of time may contribute to a malyte in which a portion of the chloride is replaced by decrease in serum potassium concentration due to increased uri- lactate or acetate, respectively.129 nary potassium excretion despite potassium administration. • Time on IV insulin infusion and dose of insulin should be mini- mized to avoid severe hypokalemia.152 7.3 | Insulin therapy • During initial volume expansion the plasma glucose concentration DKA is caused by a decrease in the effective circulating insulin level falls steeply.130 Thereafter, and after commencing insulin therapy, associated with increases in counter-regulatory hormone concentra- the plasma glucose concentration typically decreases at a rate of tions. Although rehydration alone frequently causes a marked 2 to 5 mmol/L/h, depending on the timing and amount of glucose decrease in blood glucose concentration,130,131 insulin therapy is administration134–137,139,140,153 essential to restore normal cellular metabolism, to suppress lipolysis • To prevent an unduly rapid decrease in plasma glucose concentra- and ketogenesis, and to normalize blood glucose concentrations.132 tion and hypoglycemia, 5% glucose, initially, should be added to So-called low dose IV insulin administration is safe and the IV fluid when the plasma glucose falls to approximately 14 to effective.118,133 17 mmol/L (250-300 mg/dL), or sooner if the rate of fall is precipitous. • Start insulin infusion at least 1 hour after starting fluid replace- • It may be necessary to use 10% or even 12.5% dextrose to ment therapy; that is, after the patient has received initial volume prevent hypoglycemia while continuing to infuse insulin to cor- expansion105 rect the metabolic acidosis. These glucose concentrations are • Correction of insulin deficiency often necessary to prevent hypoglycemia when insulin is • Dose: 0.05 to 0.1 unit/kg/h (eg, one method is to dilute infused at a rate of 0.1 unit/kg/h. 50 units regular [soluble] insulin in 50 mL normal saline, • If BG falls very rapidly (>5 mmol/L/h) after initial fluid expansion, 1 unit = 1 mL)134–141 consider adding glucose even before plasma glucose has • Route of administration IV decreased to 17 mmol/L (300 mg/dL). • An IV bolus should not be used at the start of therapy; it is • If biochemical parameters of DKA (venous pH, anion gap, BOHB unnecessary,140,142 may increase the risk of cerebral concentration) do not improve, reassess the patient, review WOLFSDORF ET AL. 165

insulin therapy, and consider other possible causes of impaired and inversion, ST depression, prominent U waves, apparent long response to insulin; for example, infection, errors in insulin prepa- QT interval (due to fusion of the T and U waves) indicates hypo- ration or route of administration. kalemia. Tall, peaked, symmetrical, T waves and shortening of the • In circumstances where continuous IV administration is not QT interval are signs of hyperkalemia. possible and in patients with uncomplicated DKA, hourly or • The starting potassium concentration in the infusate should be 2-hourly SC rapid-acting insulin analog (insulin lispro or insulin 40 mmol/L. Subsequent potassium replacement therapy should aspart) is safe and may be as effective as IV regular insulin be based on serum potassium measurements. infusion,153–158 but, ideally, should not be used in patients • If potassium is given with the initial rapid volume expansion, a whose peripheral circulation is impaired. Initial dose SC: 0.3 concentration of 20 mmol/L should be used. unit/kg, followed 1 hour later by SC insulin lispro or aspart at • Potassium phosphate may be used together with potassium chlo- 0.1 unit/kg every hour, or 0.15 to 0.20 units/kg every 2 to ride or acetate; for example, 20 mmol/L potassium chloride and 3 hours.158 20 mmol/L potassium phosphate or 20 mmol/L potassium phos- • If blood glucose falls to <14 mmol/L (250 mg/dL) before DKA phate and 20 mmol/L potassium acetate (C,E). Administration of has resolved, reduce SC insulin lispro or aspart to 0.05 unit/kg/h potassium entirely as potassium chloride contributes to the risk of to keep BG ≈11 mmol/L (200 mg/dL) until resolution of DKA. hyperchloremic metabolic acidosis, whereas administration • Subcutaneous administration of short-acting insulin (regular) entirely as potassium phosphate can result in hypocalcemia. every 4 hours is also a safe and effective alternative to IV insu- • Potassium replacement should continue throughout IV fluid lin infusion in children with pH ≥7.0.159 therapy. • A suggested starting dose is 0.8 to 1 unit per kg per 24-hours; • The maximum recommended rate of IV potassium replacement is the calculated 24-hour dose is divided by 6 to provide an insu- usually 0.5 mmol/kg/h. lin dose injected every 4 hours. Doses are increased or • If hypokalemia persists despite a maximum rate of potassium decreased by 10% to 20% based on the blood glucose level replacement, then the rate of insulin infusion can be reduced. before the next insulin injection. For example, if a child weighs • Profound hypokalemia (<2.5 mmol/L) in untreated DKA is rare 45 kg: 45 × 0.8 = 36 units; starting dose is 6 units.159 and necessitates vigorous potassium replacement while delaying the start of insulin therapy until serum potassium levels are 7.4 | Potassium replacement >2.5 mmol/L to reduce the risk of cardiopulmonary and neuro- muscular compromise.163 Children with DKA suffer total body potassium deficits on the order of 3to6mmol/kg.9–13 The major loss of potassium is from the intracellular 7.5 | Phosphate pool. Intracellular potassium is depleted because of transcellular shifts caused by hypertonicity (increased plasma osmolality causes solvent drag Depletion of intracellular phosphate occurs in DKA and phosphate is in which water and potassium are drawn out of cells), acidosis, and glyco- lost as a result of osmotic diuresis.5,9–11 Plasma phosphate levels fall genolysis and proteolysis secondary to insulin deficiency also cause after starting treatment and this is exacerbated by insulin, which pro- potassium efflux from cells.5 Potassium is lost from the body from vomit- motes entry of phosphate into cells.164–166 Total body phosphate ing and as a consequence of osmotic diuresis. Volume depletion causes depletion has been associated with a variety of metabolic secondary hyperaldosteronism, which promotes urinary potassium excre- disturbances.167–169 Clinically significant hypophosphatemia may tion. Total body depletion of potassium occurs; however, at presentation occur if IV therapy without food consumption is prolonged beyond serum potassium levels may be normal, increased, or decreased.160 Renal 24 hours.9–11 dysfunction, by enhancing hyperglycemia and reducing potassium excre- tion, contributes to hyperkalemia.160 Administration of insulin and the • Prospective studies involving relatively small numbers of subjects correction of acidosis drive potassium back into the cells, decreasing and with limited statistical power have not shown clinical benefit serum potassium levels.161 The serum potassium concentration may from phosphate replacement.170–175 decrease abruptly, predisposing the patient to cardiac arrhythmias. • Severe hypophosphatemia combined with phosphate depletion Replacement therapy is required regardless of the serum potas- (ie, when not solely due to intracellular phosphate translocation) sium concentration, except if renal failure is present.10,162 is uncommon, but can have severe consequences. Manifestations depend on the severity and chronicity of the phosphate depletion; • If the patient is hypokalemic, start potassium replacement at the patients usually do not have symptoms until plasma phosphate is time of initial volume expansion and before starting insulin ther- <1 mg/dL (0.32 mmol/L). apy. Otherwise, start replacing potassium after initial volume • Severe hypophosphatemia can occur during treatment of DKA; expansion and concurrent with starting insulin therapy. If the however, symptoms are uncommon because the hypophosphate- patient is hyperkalemic, defer potassium replacement therapy until mia usually is acute and there typically is no antecedent chronic urine output is documented. phosphate deficiency. • If immediate serum potassium measurements are unavailable, an • Clinical manifestations of hypophosphatemia are largely due to ECG may help to determine whether the child has hyper- or hypo- intracellular phosphate depletion. Decreased intracellular ATP kalemia.81,82 Prolongation of the PR interval, T-wave flattening levels impair cellular functions that depend on energy-rich 166 WOLFSDORF ET AL.

phosphate compounds, and a decrease in 2,3-diphosphoglycerate • Persistent ketonuria (measurement of urine ketones with test (DPG) level increases the affinity of hemoglobin for oxygen and strips is based on the nitroprusside reaction, which measures reduces oxygen release in tissues.175 Many organ systems can be acetoacetate and acetone) characteristically occurs for several affected.168,176 Manifestations include: hours after serum BOHB levels have returned to normal.68,72 • Metabolic encephalopathy (irritability, paresthesias, confusion, • Absence of ketonuria should not be used as an end-point for seizures, coma); impaired myocardial contractility and respiratory determining resolution of DKA. failure due to weakness of the diaphragm; muscle dysfunction • When ketoacidosis has resolved, oral intake is tolerated, and the with proximal myopathy, dysphagia and ileus; rare hematologic change to SC insulin is planned, a dose of basal (long- or effects include hemolysis, decreased phagocytosis and granulo- intermediate-acting) insulin should be administered in addition to 158 cyte chemotaxis, defective clot retraction, and thrombocytopenia. rapid- or short-acting insulin. The most convenient time to change Acute hypophosphatemia in a patient with preexisting severe to SC insulin is just before a mealtime. There may, also, be benefits to earlier administration of a dose of basal insulin while the patient is phosphate depletion can lead to rhabdomyolysis.168,177,178 still receiving IV insulin infusion. For example, in one uncontrolled • Severe hypophosphatemia associated with any of the above study, 0.3 units/kg of SC insulin glargine was given in the first 6 hours symptoms should be treated.179,180 of management and led to faster resolution of DKA.192 Another ret- • Administration of phosphate may induce hypocalcemia.181,182 rospective uncontrolled study showed that co-administration of • Potassium phosphate salts may be safely used as an alternative to insulin glargine (approximately 0.4 units/kg) early in the course of or combined with potassium chloride or acetate, provided that DKA treatment was well tolerated, did not increase the risk of hypo- careful monitoring of serum calcium is performed to avoid hypo- glycemia, but was associated with more frequent hypokalemia.193 calcemia (C).181,182 • To prevent rebound hyperglycemia the first SC injection should be given 15 to 30 minutes (with rapid-acting insulin) or 1 to 7.6 | Acidosis 2 hours (with regular insulin) before stopping the insulin infusion to allow sufficient time for the insulin to be absorbed. With Severe acidosis is reversible by fluid and insulin replacement; insulin intermediate- or long-acting insulin, the overlap should be longer stops further ketoacid production and allows ketoacids to be metabo- and the rate of IV insulin administration gradually decreased. For lized, which generates bicarbonate. Treatment of hypovolemia example, for patients on a basal-bolus insulin regimen, the first improves tissue perfusion and renal function, thereby increasing the dose of basal insulin may be administered in the evening and the excretion of organic acids. IV insulin infusion is stopped the next morning. Controlled trials have shown no clinical benefit from bicarbonate • The regimen, dose and type of SC insulin should be according to administration.183–186 Bicarbonate therapy may cause paradoxical CNS local preferences and circumstances. acidosis187,188 and rapid correction of acidosis with bicarbonate causes • After transitioning to SC insulin, frequent blood glucose monitoring is 187,189,190 hypokalemia. Bicarbonate administration may be beneficial in required to avoid marked hyperglycemia and hypoglycemia. the rare patient with life-threatening hyperkalemia or unusually severe acidosis (vpH <6.9) that has compromised cardiac contractility.191

8.1 | Morbidity and mortality • If bicarbonate is considered necessary, cautiously give 1 to In population studies, the mortality rate from DKA in children is 0.15% 2 mmol/kg over 60 minutes. to 0.30%194–196 and may be decreasing.196,197 The Centers for Disease Control and Prevention (CDC) from US National Vital Statistics System Complications of therapy data 1968 to 2009 found that mortality has decreased from an annual rate of 2.69 per million for the period 1968 to 1969 to a rate of 1.05 • Cerebral edema per million in 2008 to 2009.197 A number of possible reasons have • Hypokalemia been proposed for the reduction in diabetes-related deaths in children, • Hyperchloremic acidosis including improved diabetes care and treatment, increased awareness • Hypoglycemia of diabetes symptoms, possibly resulting in earlier recognition and • Inadequate rehydration treatment, and advances in education regarding diabetes and manage- ment of DKA. However, recent data show that DKA is still the leading cause of death in subjects with T1D diagnosed less than 15 years of age198 and mortality risk is substantially increased in patients with chronically poor glycemic control and recurrent DKA.199,200 8 | INTRODUCTION OF ORAL FLUIDS AND Cerebral injury is the major cause of mortality and morbidity195,201 and TRANSITION TO SC INSULIN INJECTIONS cerebral edema accounts for 60% to 90% of all DKA deaths.102,202 From 10% to 25% of survivors of cerebral edema have significant residual morbid- • Oral fluids should be introduced only when substantial clinical ity.102,202,203 Children without overt neurological symptoms during DKA improvement has occurred (mild acidosis/ketosis may still be present). treatment may have subtle evidence of brain injury, particularly memory WOLFSDORF ET AL. 167 deficits, after recovery from DKA.204 Recent studies have shown that injury,102,227–230 which have led to the formulation of an alternative acutely impaired cognition (“mental state”) is common at presentation with hypothesis; namely, that factors intrinsic to DKA may be the cause of newly diagnosed type 1 diabetes and more likely to occur in children who brain injury, which could be worsened during treatment.231,232 It is present with DKA. Impaired cognition at presentation also is associated with noteworthy that the degree of cerebral edema that develops during poorer attention and memory in the week following diagnosis and lower DKA correlates with the degree of dehydration and hyperventilation intelligence quotient at 6 months compared to children with unimpaired at presentation, but not with initial osmolality or osmotic changes dur- cognition at diagnosis.205 Magnetic resonance imaging (MRI), spectroscopy, ing treatment.218 Disruption of the blood-brain-barrier has been and cognitive assessments at diagnosis and up to 6 months postdiagnosis found in cases of fatal cerebral edema associated with DKA,233,234 show morphologic and functional brain changes that are associated with which further supports the view that cerebral edema is not simply 206 adverse neurocognitive outcomes in the medium term. Furthermore, caused by a reduction in serum osmolality. DKA at onset of type 1 diabetes predicts worse long-term glycemic control Demographic factors associated with an increased risk of cerebral 207,208 independent of demographic and socioeconomic factors. edema include: Other rare causes of morbidity and mortality include:

• Younger age235 • * Hypokalemia • New onset diabetes195,235 • Hypocalcemia, hypomagnesemia • Longer duration of symptoms236 • Severe hypophosphatemia* • Hypochloremic alkalosis209 These risk associations may reflect the greater likelihood of • Hypoglycemia severe DKA. • Other central nervous system complications include dural sinus Epidemiological studies have identified several potential risk fac- thrombosis, basilar artery thrombosis, intracranial hemorrhage, tors at diagnosis or during treatment of DKA. These include: cerebral infarction210–212 • Venous thrombosis83,84* • Greater hypocapnia at presentation after adjusting for degree of • Pulmonary embolism* acidosis102,218,237 • Sepsis • Increased serum urea nitrogen at presentation102,218 • Rhinocerebral or pulmonary mucormycosis213 • More severe acidosis at presentation105,238,239 • Aspiration pneumonia* • Bicarbonate treatment for correction of acidosis102,240 • Pulmonary edema* • A marked early decrease in serum effective osmolality119,239 • Adult respiratory distress syndrome (ARDS) • An attenuated rise in serum sodium concentration or an early fall • Pneumothorax, pneumomediastinum and subcutaneous emphysema214 in glucose-corrected sodium during therapy100–102,239 • Rhabdomyolysis* • Greater volumes of fluid given in the first 4 hours105,237,239 • Ischemic bowel necrosis • Administration of insulin in the first hour of fluid treatment105 • Acute kidney injury including renal failure215* • Acute pancreatitis216*

*These complications, often with fatality, have been more fre- Signs and symptoms of cerebral edema quent in HHS (see Reference 41). The pathophysiology and manage- include: ment of HHS are discussed below. • Onset of headache after beginning treatment or pro- 8.2 | Cerebral edema gressively worsening headache. • Change in neurological status (irritability, confusion, The incidence of clinically overt cerebral edema in national population inability to arouse, incontinence). studies is 0.5% to 0.9% and the mortality rate is 21% to • Specific neurological signs (eg, cranial nerve palsies, 24%.102,202,203 Mental status abnormalities (GCS scores <14), how- papilledema). ever, occur in approximately 4% to 15% of children treated for DKA • Cushing's triad (rising blood pressure, bradycardia, and and are associated with evidence of cerebral edema on neuroimag- respiratory depression) is a late but important sign of ing.217,218 Cerebral edema is rarely seen after adolescence. Neuroim- increased intracranial pressure. aging studies have led to the appreciation that cerebral edema is not a • Decreased O saturation. rare phenomenon in children with DKA, but occurs frequently and 2 with varying severity.217,219,220 Clinically overt cerebral edema repre- sents the most severe manifestation of a common phenomenon.221 The cause of cerebral edema is controversial. Some have Clinically significant cerebral edema usually develops within the explained the pathogenesis as the result of rapid fluid administration first 12 hours after treatment has started but can occur before – with abrupt changes in serum osmolality.120,222–226 More recent treatment has begun102,203,241 244 or, rarely, may develop as late as investigations, however, have found that dehydration and cerebral 24 to 48 hours after the start of treatment.102,235,245 Symptoms and hypoperfusion may be associated with DKA-related cerebral signs are variable. Although mild to moderate headache at 168 WOLFSDORF ET AL. presentation may not be unusual, development of a severe head- • Hypertonic saline (3%) 2.5 mL/kg is equimolar to mannitol ache after commencing treatment is always concerning. A method 0.5 g/kg. of clinical diagnosis based on bedside evaluation of neurological • A recent 11-year retrospective cohort study showed that state is shown below.246 One diagnostic criterion, two major criteria, hypertonic saline has replaced mannitol as the most commonly or one major and two minor criteria have a sensitivity of 92% and a used hyperosmolar agent in many US institutions. Although false positive rate of only 4%. Signs that occur before treatment controversial and further investigation is needed, the data sug- should not be considered in the diagnosis of cerebral edema. Neuro- gest that hypertonic saline may not have benefits over manni- imaging is not required for diagnosis of cerebral edema. tol and may be associated with a higher mortality rate.196,252 Diagnostic criteria • Elevate the head of the bed to 30 and keep the head in the mid- line position. • Abnormal motor or verbal response to pain • Intubation may be necessary for the patient with impending respi- • Decorticate or decerebrate posture ratory failure due to severe neurologic compromise. • Cranial nerve palsy (especially III, IV, and VI) • After treatment for cerebral edema has been started, cranial imag- • Abnormal neurogenic respiratory pattern (eg, grunting, tachypnea, ing may be considered as with any critically ill patient with Cheyne-Stokes respiration, apneusis) encephalopathy or acute focal neurologic deficit. However, treat- ment of the clinically symptomatic patient should not be delayed Major criteria in order to obtain imaging.253 The primary concern that would warrant neuroimaging is whether the patient has a lesion requir- • Altered mentation, confusion, fluctuating level of consciousness ing emergency neurosurgery (eg, intracranial hemorrhage) or a • Sustained heart rate deceleration (decrease more than 20 beats lesion that may necessitate anticoagulation (eg, cerebrovascular per minute) not attributable to improved intravascular volume or thrombosis), as suggested by clinical findings of focal or severe, sleep state progressive headache, or focal neurologic deficit.211,254–256 • Age-inappropriate incontinence

Minor criteria 8.4 | Hyperglycemic hyperosmolar state This syndrome is characterized by extremely elevated serum glucose • Vomiting concentrations and hyperosmolality without significant ketosis.41 The • Headache incidence of HHS is increasing37,38,257 and a recent study found HHS • Lethargy or not easily arousable in 2% of youth with type 2 diabetes at presentation34; nonetheless, it • Diastolic blood pressure >90 mm Hg is considerably less frequent in children and adolescents than DKA. • Age <5 years Unlike the usual symptoms of DKA (hyperventilation, vomiting, and abdominal pain), which typically bring children to medical attention, the A chart with the reference ranges for blood pressure and heart gradually increasing polyuria and polydipsia of HHS may go unrecog- rate (which vary depending on height, weight, and gender) should be nized resulting in profound dehydration and electrolyte losses at the readily available, either in the patient's chart or at the bedside. time of presentation. In adults, fluid losses in HHS have been estimated The appearance of diabetes insipidus, manifested by increased to be twice those of DKA; furthermore, obesity and hyperosmolality urine output with a concomitant marked increase in the serum sodium can make the clinical assessment of dehydration challenging. Despite concentration, reflecting loss of free water in the urine, is a sign of cere- severe volume depletion and electrolyte losses, hypertonicity preserves bral herniation causing interruption of blood flow to the pituitary gland. intravascular volume and signs of dehydration may be less evident. During therapy, decreasing serum osmolality (from enhanced glu- 8.3 | Treatment of cerebral edema cosuria and insulin-mediated glucose uptake) results in movement of water out of the intravascular space resulting in decreased intravascu- • Initiate treatment as soon as the condition is suspected. lar volume, and pronounced osmotic diuresis may continue for many • Adjust fluid administration rate as needed to maintain normal hours in patients with extremely increased plasma glucose concentra- blood pressure while avoiding excessive fluid administration that tions. Early in the course of treatment, urinary fluid losses may be con- might increase cerebral edema formation. Assiduously avoid siderable and because intravascular volume may decrease rapidly hypotension that might compromise cerebral perfusion pressure. during treatment in patients with HHS, more aggressive replacement • Hyperosmolar agents should be readily available at the bedside. of intravascular volume (as compared to treatment of children with • Give mannitol, 0.5 to 1 g/kg IV over 10 to 15 minutes.247–249 The DKA) is required to avoid vascular collapse. effect of mannitol should be apparent after ~15 minutes, and is expected to last about 120 minutes. If necessary, the dose can be 8.5 | Treatment of HHS repeated after 30 minutes. • Hypertonic saline (3%), suggested dose 2.5 to 5 mL/kg over 10 to There are no prospective data to guide treatment of children and ado- 15 minutes, may be used as an alternative to mannitol, or in addi- lescents with HHS. The following recommendations are based on tion to mannitol if there has been no response to mannitol within extensive experience in adults and an appreciation of the pathophysi- 15 to 30 minutes.250,251 ological differences between HHS and DKA41; see Figure 3. Patients WOLFSDORF ET AL. 169

Fluids Electrolytes Insulin

HHS Hyperosmolar DKA

Bolus 0.9%saline 20 cc/kg, repeat When serum K+ <5 mEq/L, Start insulin infusion when BG Start insulin after initial until perfusion established start replacement with K no longer decreases with fluid fluid bolus 40 mEq/L alone

Maintenance fluids plus Replace urine output IV regular insulin IV regular insulin deficit replacement 0.025‐0.05 unit/kg/hr 0.05‐0.1 unit/kg/hr over 24‐48 hours; 0.45‐ Monitor electrolytes, depending on degree 0.75% saline calcium, magnesium, of acidosis phosphate every 2‐4 hours

Titrate insulin dose to decreaseblood glucose 4‐ 5.5 mmol/L (75-100 mg/dL) per hour

Frequently assess circulatory status Adjust rate and electrolyte composition of fluids as needed

FIGURE 3 Treatment of hyperglycemic hyperosmolar syndrome (HHS) (from Reference 41)

should be admitted to an intensive care unit or comparable setting been associated with failure of the corrected serum sodium con- where expert medical, nursing and laboratory services are available. centration to decline with treatment, which may be an indication for hemodialysis. Hemodialysis has resulted in 80% survival in 8.6 | Fluid therapy contrast to 20% with peritoneal dialysis.38

The goal of initial fluid therapy is to expand the intra- and • Although there are no data to indicate an optimal rate of extravascular volume and restore normal renal perfusion. The rate of decline in serum sodium concentration, 0.5 mmol/L/h has been fluid replacement should be more rapid than is recommended recommended for hypernatremic dehydration.258 With ade- for DKA. quate rehydration alone (ie, before commencing insulin ther- apy), serum glucose concentrations should decrease by 4.1 to • The initial bolus should be ≥20 mL/kg of isotonic saline (0.9% 5.5 mmol/L (75-100 mg/dL) per hour.259,260 NaCl) and a fluid deficit of approximately 12% to 15% of • A more rapid rate of decline in serum glucose concentration is body weight should be assumed. Additional fluid boluses typical during the first several hours of treatment when an should be given rapidly, if necessary, to restore peripheral expanded vascular volume leads to improved renal perfusion. perfusion. If there is a continued rapid fall in serum glucose (>5 mmol/L, • Thereafter, 0.45% to 0.75% NaCl should be administered to 90 mg/dL/h) after the first few hours, consider adding 2.5% or replace the deficit over 24 to 48 hours. 5% glucose to the rehydration fluid. Failure of the expected • The goal is to promote a gradual decline in corrected serum decrease of plasma glucose concentration should prompt reas- sodium concentration and serum osmolality. • Because isotonic fluids are more effective in maintaining circula- sessment and evaluation of renal function. • Unlike treatment of DKA, replacement of urinary losses is tory volume, isotonic saline should be restarted if perfusion and 141 hemodynamic status appear inadequate as serum osmolality recommended. The typical urine sodium concentration dur- declines. ing an osmotic diuresis approximates 0.45% saline; however, • Serum sodium concentrations should be measured frequently and when there is concern about the adequacy of circulatory vol- the sodium concentration in fluids adjusted to promote a gradual ume, urinary losses may be replaced with a fluid containing a decline in corrected serum sodium concentration. Mortality has higher sodium concentration. 170 WOLFSDORF ET AL.

8.7 | Insulin therapy • Replacement of magnesium should be considered in the occa- sional patient who experiences severe hypomagnesemia and Whereas tissue hypoperfusion in HHS commonly causes lactic acido- hypocalcemia during therapy. The recommended dose is 25 to sis, ketosis usually is minimal. Early insulin administration is unneces- 50 mg/kg per dose for 3 to 4 doses given every 4 to 6 hours sary in HHS. Fluid administration alone causes a marked decline in with a maximum infusion rate of 150 mg/min and 2 g/h. serum glucose concentration as a result of dilution, improved renal perfusion leading to increased glucosuria, and increased tissue glucose uptake with improved circulation. The osmotic pressure exerted by glucose within the vascular space contributes to the maintenance of 8.9 | Complications blood volume. A rapid fall in serum glucose concentration and osmo- • Venous thrombosis associated with use of central venous cathe- lality after insulin administration may lead to circulatory compromise ters is a common hazard in HHS.83 Prophylactic use of low-dose and venous thrombosis unless fluid replacement is adequate. Patients heparin has been suggested in adults but there are no data to with HHS also have extreme potassium deficits; a rapid insulin- indicate benefit from this practice. Heparin treatment should be induced shift of potassium to the intracellular space can trigger an reserved for children who require central venous catheters for arrhythmia. physiologic monitoring or venous access and are immobile for more than 24 to 48 hours.41 The central venous catheter should • Insulin administration should be initiated when serum glucose not be used for insulin administration because the large dead concentration is no longer declining at a rate of at least 3 mmol/L space may cause erratic insulin delivery. (~50 mg/dL) per hour with fluid administration alone. • Rhabdomyolysis may occur in children with HHS resulting in • In patients with more severe ketosis and acidosis, however, insu- acute kidney failure, severe hyperkalemia, hypocalcemia, and lin administration should be initiated earlier. muscle swelling causing compartment syndrome.261 The classic • Continuous administration of insulin at a rate of 0.025 to symptom triad of rhabdomyolysis includes myalgia, weakness, and 0.05 units/kg/h can be used initially, with the dosage titrated to dark urine; monitoring creatine kinase concentrations every 2 to achieve a decrease in serum glucose concentration of 3 to 3 hours is recommended for early detection. 4 mmol/L (~50-75 mg/dL) per hour. • For unknown reasons, several children with HHS have had clinical • Insulin boluses are not recommended. manifestations consistent with malignant hyperthermia, which is associated with a high mortality rate.36,262–264 Patients who have 8.8 | Electrolytes a fever associated with a rise in creatine kinase concentrations

In general, deficits of potassium, phosphate, and magnesium are may be treated with dantrolene, which reduces release of calcium greater in HHS than DKA. from the sarcoplasmic reticulum and stabilizes calcium metabolism within muscle cells. Nonetheless, of the three reported patients 262,264 • Potassium replacement (40 mmol/L of replacement fluid) should with HHS treated with dantrolene only one survived. • begin as soon as serum potassium concentration is within the nor- Altered mental status is common in adults whose serum osmolal- 38 mal range and adequate renal function has been established. ity exceeds 330 mOsm/kg; however, cerebral edema is rare. • Higher rates of potassium administration may be necessary Among 96 cases of HHS reported in the literature to 2010, after starting an insulin infusion including 32 deaths, there was only one instance of cerebral • Serum potassium concentrations should be monitored every edema.38 A decline in mental status after hyperosmolality has 2 to 3 hours along with ECG monitoring improved with treatment is unusual and should be promptly • Hourly potassium measurements may be necessary if the investigated. patient has hypokalemia • Bicarbonate therapy is contraindicated; it increases the risk of hypokalemia and may adversely affect tissue oxygen delivery. 8.10 | Mixed HHS and DKA • Severe hypophosphatemia may lead to rhabdomyolysis, hemolytic uremia, muscle weakness and paralysis. Although administration Treatment must take into account potential complications of both of phosphate is associated with a risk of hypocalcemia, an IV solu- DKA and HHS. Mental status must be closely monitored and frequent tion that contains a 50:50 mixture of potassium phosphate and reassessment of circulatory status and fluid balance is necessary to another suitable potassium salt (potassium chloride or potassium guide therapy. To maintain an adequate circulatory volume, the rate acetate), generally permits adequate phosphate replacement of fluid and electrolyte administration usually exceeds that required while avoiding clinically significant hypocalcemia. for the typical case of DKA. Insulin is necessary to resolve ketosis and • Serum phosphate concentrations should be measured every arrest hepatic gluconeogenesis; however, insulin infusion should be 3 to 4 hours. deferred until after the patient has received an initial fluid bolus and • Patients with HHS frequently have large magnesium deficits, but the circulation has been stabilized. Serum potassium and phosphate there are no data to determine whether replacement of magne- concentrations should be carefully monitored as described above sium is beneficial. for HHS. WOLFSDORF ET AL. 171

8.11 | Prevention of recurrent DKA ‡Some experts use 1.6 rather than 2 as the coefficient in the equation.92 Management of an episode of DKA is not complete until its cause has been identified and an attempt made to treat it. CONFLICT OF INTEREST • Insulin omission, either inadvertently or deliberately, is the cause in most cases. J.W. receives royalties as a section editor for UpToDate. M.F. is mem- • The most common cause of DKA in insulin pump users is failure ber of Lilly Diabetes, Eli Lilly Ges.m.b. (advisory board); Medtronic to take extra insulin with a pen or syringe when hyperglycemia (speaker fee); Ferring pharmaeuticals Ges m.b.h. (prize for best scien- and hyperketonemia or ketonuria occur. tific presentation). R.H. receives lecture honoraria from Novo Nordisk, • Home measurement of blood BOHB concentrations, when com- Medtronic; Advisory Board Novo Nordisk, Medtronic, Insulet; Data pared to urine ketone testing, decreases diabetes-related hospital Monitoring Committee Astra Zeneca. M.A.S. was a member of a Sci- visits (both emergency department visits and hospitalizations) by entific Advisory Board for NovoNordisk during 2017 and 2018. A.R., 265 the early identification and treatment of ketosis. Blood BOHB N.G., M.A. and E.C. have no conflicts of interest to declare. measurements may be especially valuable to prevent DKA in patients who use a pump because interrupted insulin delivery rap- ORCID idly leads to ketosis. Maria Fritsch http://orcid.org/0000-0001-5304-673X • There may be dissociation between urine ketone (sodium Ethel Codner http://orcid.org/0000-0002-2899-2705 nitroprusside only measures acetoacetate and acetone) and serum BOHB concentrations, which may be increased to levels consistent with DKA at a time when a urine ketone test is neg- REFERENCES ative or shows only trace or small ketonuria.266 1. Foster DW, McGarry JD. The metabolic derangements and treatment of diabetic ketoacidosis. N Engl J Med. 1983;309(3):159-169. • There usually is an important psychosocial reason for insulin 2. Kitabchi AE, Umpierrez GE, Murphy MB, Kreisberg RA. Hyperglyce- omission. mic crises in adult patients with diabetes: a consensus statement • an attempt to lose weight in an adolescent girl with an eating from the American Diabetes Association. Diabetes Care. 2006;29(12): disorder, 2739-2748. 3. Hanas R, Lindgren F, Lindblad B. A 2-yr national population study of • a means of escaping an intolerable or abusive home situation, pediatric ketoacidosis in Sweden: predisposing conditions and insulin • depression or other reason for inability of the patient to man- pump use. Pediatr Diabetes. 2009;10(1):33-37. age the diabetes unassisted. 4. Cox K, Cocchi MN, Salciccioli JD, Carney E, Howell M, Donnino MW. Prevalence and significance of lactic acidosis in diabetic ketoacidosis. • A psychiatric social worker or clinical psychologist should be con- J Crit Care. 2012;27(2):132-137. sulted to identify the psychosocial reason(s) contributing to devel- 5. Palmer BF, Clegg DJ. Electrolyte and acid-base disturbances in opment of DKA. patients with diabetes mellitus. N Engl J Med. 2015;373(6):548-559. • An infection is rarely the cause of DKA when the patient/family is 6. Deeter KH, Roberts JS, Bradford H, et al. Hypertension despite dehy- dration during severe pediatric diabetic ketoacidosis. Pediatr Diabe- properly educated in diabetes management and is receiving tes. 2011;12(4 pt 1):295-301. appropriate follow-up care by a diabetes team with a 24-hour 7. McDonnell CM, Pedreira CC, Vadamalayan B, Cameron FJ, telephone helpline.267–269 Werther GA. Diabetic ketoacidosis, hyperosmolarity and hypernatre- mia: are high-carbohydrate drinks worsening initial presentation? • Insulin omission can be prevented by comprehensive programs Pediatr Diabetes. 2005;6(2):90-94. that provide education, psychosocial evaluation and treatment 8. Carlotti AP, St George-Hyslop C, Guerguerian AM, Bohn D, combined with adult supervision of the entire process of insulin Kamel KS, Halperin M. Occult risk factor for the development of cerebral edema in children with diabetic ketoacidosis: possible role administration.270 for stomach emptying. Pediatr Diabetes. 2009;10(8):522-533. • Parents and patients should learn how to recognize and treat 9. Atchley D, Loeb R, Richards D Jr, Benedict E, Driscoll M. On diabetic ketosis and impending DKA with additional rapid- or short- ketoacidosis: a detailed study of electrolyte balances following the withdrawal and reestablishment of insulin therapy. J Clin Invest. acting insulin and oral fluids 1933;12:297-326. • Families should have access to a 24-hour telephone helpline 10. Nabarro J, Spencer A, Stowers J. Metabolic studies in severe diabetic for emergency advice and treatment267 ketosis. Q J Med. 1952;82:225-248. • When a responsible adult administers insulin there may be as 11. Butler A, Talbot N, Burnett C, Stanbury J, MacLachlan E. Metabolic studies in diabetic coma. Trans Assoc Am Phys. 1947;60:102-109. much as a 10-fold reduction in frequency of recurrent DKA.270 12. Danowski T, Peters J, Rathbun J, Quashnock J, Greenman L. Studies in diabetic acidosis and coma, with particular emphasis on the reten- tion of administered potassium. J Clin Invest. 1949;28:1-9. ENDNOTES 13. Darrow D, Pratt E. Retention of water and electrolyte during recov- ery in a patient with diabetic acidosis. J Pediatr. 1952;41:688-696. 14. Holliday MA, Segar WE. The maintenance need for water in paren- *Nitroprusside reaction method. † teral fluid therapy. Pediatrics. 1957;19(5):823-832. Serum bicarbonate concentration alone can substitute for vpH to 15. Friedman AL. Pediatric hydration therapy: historical review and a diagnose DKA and classify severity in children with new onset diabe- new approach. Kidney Int. 2005;67(1):380-388. tes mellitus. It is suggested as an alternative to reliance on vpH, espe- 16. Collier S, Gura K, deLoid L, Dalton M. L S. Parenteral Nutrition. In: cially in settings in which access to vpH measurement is limited.26 Sonneville K, Duggan C, eds. Manual of Pediatric Nutrition. 5th 172 WOLFSDORF ET AL.

ed. Shelton, CT: People's Medical Publishing House-USA; 2014: 39. Bagdure D, Rewers A, Campagna E, Sills MR. Epidemiology of hyper- 196-248. glycemic hyperosmolar syndrome in children hospitalized in USA. 17. Darrow DC. The physiologic basis for estimating requirements for Pediatr Diabetes. 2013;14(1):18-24. parenteral fluids. Pediatr Clin N Am. 1959;6(1):29-41. 40. Temple IK, Shield JP. 6q24 transient neonatal diabetes. Rev Endocr 18. Dunger DB, Sperling MA, Acerini CL, et al. ESPE/LWPES consensus Metab Disord. 2010;11(3):199-204. statement on diabetic ketoacidosis in children and adolescents. Arch 41. Zeitler P, Haqq A, Rosenbloom A, Glaser N. Hyperglycemic hyperos- Dis Child. 2004;89(2):188-194. molar syndrome in children: pathophysiological considerations and 19. Sheikh-Ali M, Karon BS, Basu A, et al. Can serum suggested guidelines for treatment. J Pediatr. 2011;158(1):9-14, beta-hydroxybutyrate be used to diagnose diabetic ketoacidosis? 14.e1-2. Diabetes Care. 2008;31(4):643-647. 42. Levy-Marchal C, Papoz L, de Beaufort C, et al. Clinical and laboratory 20. Burge MR, Hardy KJ, Schade DS. Short-term fasting is a mechanism features of type 1 diabetic children at the time of diagnosis. Diabet for the development of euglycemic ketoacidosis during periods of Med. 1992;9(3):279-284. insulin deficiency. J Clin Endocrinol Metab. 1993;76(5):1192-1198. 43. Komulainen J, Lounamaa R, Knip M, Kaprio EA, Akerblom HK. Ketoa- 21. Pinkney JH, Bingley PJ, Sawtell PA, Dunger DB, Gale EA. Presenta- cidosis at the diagnosis of type 1 (insulin dependent) diabetes melli- tus is related to poor residual beta cell function. Childhood Diabetes tion and progress of childhood diabetes mellitus: a prospective in Finland Study Group. Arch Dis Child. 1996;75(5):410-415. population-based study. The Bart's-Oxford Study Group. Diabetolo- 44. Levy-Marchal C, Patterson CC, Green A. Geographical variation of gia. 1994;37(1):70-74. presentation at diagnosis of type I diabetes in children: the EURO- 22. Bas VN, Uytun S, Torun YA. Diabetic euglycemic ketoacidosis in DIAB study. European and Dibetes. Diabetologia. 2001;44((suppl 3)): newly diagnosed type 1 diabetes mellitus during Ramadan fasting. J B75-B80. Pediatr Endocrinol Metab. 2015;28(3–4):333-335. 45. Hanas R, Lindgren F, Lindblad B. Diabetic ketoacidosis and cerebral 23. Peters AL, Buschur EO, Buse JB, Cohan P, Diner JC, Hirsch IB. Eugly- oedema in Sweden – a 2-year paediatric population study. Diabet cemic diabetic ketoacidosis: a potential complication of treatment Med. 2007;24(10):1080-1085. with sodium-glucose cotransporter 2 inhibition. Diabetes Care. 2015; 46. Rodacki M, Pereira JR, Nabuco de Oliveira AM, et al. Ethnicity and 38(9):1687-1693. young age influence the frequency of diabetic ketoacidosis at the 24. Goldenberg RM, Berard LD, Cheng AY, et al. SGLT2 onset of type 1 diabetes. Diabetes Res Clin Pract. 2007;78(2): inhibitor-associated diabetic ketoacidosis: clinical review and recom- 259-262. mendations for prevention and diagnosis. Clin Ther. 2016;38(12): 47. Usher-Smith JA, Thompson M, Ercole A, Walter FM. Variation 2654-2664.e1. between countries in the frequency of diabetic ketoacidosis at first 25. Misaghian-Xanthos N, Shariff AI, Mekala K, et al. Sodium-glucose presentation of type 1 diabetes in children: a systematic review. Dia- cotransporter 2 inhibitors and diabetic ketoacidosis: a case series betologia. 2012;55(11):2878-2894. from three academic institutions. Diabetes Care. 2017;40(6):e65-e66. 48. Fritsch M, Schober E, Rami-Merhar B, Hofer S, Frohlich-Reiterer E, 26. von Oettingen J, Wolfsdorf J, Feldman HA, Rhodes ET. Use of serum Waldhoer T. Diabetic ketoacidosis at diagnosis in Austrian children: a bicarbonate to substitute for venous pH in new-onset diabetes. Pedi- population-based analysis, 1989-2011. J Pediatr. 2013;163: atrics. 2015;136(2):e371-e377. 1484-8.e1. 27. Mayer-Davis EJ, Lawrence JM, Dabelea D, et al. Incidence trends of 49. Zucchini S, Scaramuzza AE, Bonfanti R, et al. A multicenter retrospec- type 1 and type 2 diabetes among youths, 2002-2012. N Engl J Med. tive survey regarding diabetic ketoacidosis management in Italian 2017;376(15):1419-1429. children with type 1 diabetes. J Diabetes Res. 2016;2016:5719470. 28. Ingelfinger JR, Jarcho JA. Increase in the incidence of diabetes and its 50. Letourneau LR, Carmody D, Wroblewski K, et al. Diabetes presenta- implications. N Engl J Med. 2017;376(15):1473-1474. tion in infancy: high risk of diabetic ketoacidosis. Diabetes Care. 29. Fazeli Farsani S, van der Aa MP, van der Vorst MM, Knibbe CA, de 2017;40(10):e147-e148. Boer A. Global trends in the incidence and prevalence of type 2 dia- 51. Quinn M, Fleischman A, Rosner B, Nigrin DJ, Wolfsdorf JI. Character- betes in children and adolescents: a systematic review and evaluation istics at diagnosis of type 1 diabetes in children younger than 6 years. of methodological approaches. Diabetologia. 2013;56(7):1471-1488. J Pediatr. 2006;148(3):366-371. 30. Dabelea D, Rewers A, Stafford JM, et al. Trends in the prevalence of 52. Bui H, To T, Stein R, Fung K, Daneman D. Is diabetic ketoacidosis at ketoacidosis at diabetes diagnosis: the SEARCH for diabetes in youth disease onset a result of missed diagnosis? J Pediatr. 2010;156(3): study. Pediatrics. 2014;133(4):e938-e945. 472-477. 31. American Diabetes Association. Type 2 diabetes in children and ado- 53. Szypowska A, Skorka A. The risk factors of ketoacidosis in children with newly diagnosed type 1 diabetes mellitus. Pediatr Diabetes. lescents. Pediatrics. 2000;105((3, pt 1)):671-680. 2011;12(4, pt 1):302-306. 32. Rewers A, Klingensmith G, Davis C, et al. Presence of diabetic ketoa- 54. Komulainen J, Kulmala P, Savola K, et al. Clinical, autoimmune, and cidosis at diagnosis of diabetes mellitus in youth: the Search for Dia- genetic characteristics of very young children with type 1 diabetes. betes in Youth Study. Pediatrics. 2008;121(5):e1258-e1266. Childhood Diabetes in Finland (DiMe) Study Group. Diabetes Care. 33. Gungor N, Hannon T, Libman I, Bacha F, Arslanian S. Type 2 diabetes 1999;22(12):1950-1955. mellitus in youth: the complete picture to date. Pediatr Clin N Am. 55. Usher-Smith JA, Thompson MJ, Sharp SJ, Walter FM. Factors associ- 2005;52(6):1579-1609. ated with the presence of diabetic ketoacidosis at diagnosis of diabe- 34. Klingensmith GJ, Connor CG, Ruedy KJ, et al. Presentation of youth tes in children and young adults: a systematic review. BMJ. 2011; with type 2 diabetes in the Pediatric Diabetes Consortium. Pediatr 343:d4092. Diabetes. 2016;17(4):266-273. 56. Rosilio M, Cotton JB, Wieliczko MC, et al. Factors associated with 35. Chase HP, Garg SK, Jelley DH. Diabetic ketoacidosis in children and glycemic control. A cross-sectional nationwide study in 2,579 French the role of outpatient management. Pediatr Rev. 1990;11(10): children with type 1 diabetes. The French Pediatric Diabetes Group. 297-304. Diabetes Care. 1998;21(7):1146-1153. 36. Morales AE, Rosenbloom AL. Death caused by hyperglycemic hyper- 57. Smith CP, Firth D, Bennett S, Howard C, Chisholm P. Ketoacidosis osmolar state at the onset of type 2 diabetes. J Pediatr. 2004;144(2): occurring in newly diagnosed and established diabetic children. Acta 270-273. Paediatr. 1998;87(5):537-541. 37. Canarie MF, Bogue CW, Banasiak KJ, Weinzimer SA, 58. Morris AD, Boyle DI, McMahon AD, Greene SA, MacDonald TM, Tamborlane WV. Decompensated hyperglycemic hyperosmolarity Newton RW. Adherence to insulin treatment, glycaemic control, and without significant ketoacidosis in the adolescent and young adult ketoacidosis in insulin-dependent diabetes mellitus. The DARTS/ population. J Pediatr Endocrinol Metab. 2007;20(10):1115-1124. MEMO Collaboration. Diabetes Audit and Research in Tayside Scot- 38. Rosenbloom AL. Hyperglycemic hyperosmolar state: an emerging land Medicines Monitoring Unit. Lancet. 1997;350(9090): pediatric problem. J Pediatr. 2010;156(2):180-184. 1505-1510. WOLFSDORF ET AL. 173

59. Rewers A, Chase HP, Mackenzie T, et al. Predictors of acute compli- 80. Davids MR, Edoute Y, Stock S, Halperin ML. Severe degree of hyper- cations in children with type 1 diabetes. JAMA. 2002;287(19): glycaemia: insights from integrative physiology. QJM. 2002;95(2): 2511-2518. 113-124. 60. Cengiz E, Xing D, Wong JC, et al. Severe hypoglycemia and diabetic 81. Malone JI, Brodsky SJ. The value of electrocardiogram monitoring in ketoacidosis among youth with type 1 diabetes in the T1D exchange diabetic ketoacidosis. Diabetes Care. 1980;3(4):543-547. clinic registry. Pediatr Diabetes. 2013;14(6):447-454. 82. Soler NG, Bennett MA, Fitzgerald MG, Malins JM. Electrocardiogram 61. Maahs DM, Hermann JM, Holman N, et al. Rates of diabetic ketoaci- as a guide to potassium replacement in diabetic ketoacidosis. Diabe- dosis: international comparison with 49,859 pediatric patients with tes. 1974;23(7):610-615. type 1 diabetes from England, Wales, the U.S., Austria, and Germany. 83. Gutierrez JA, Bagatell R, Samson MP, Theodorou AA, Berg RA. Fem- Diabetes Care. 2015;38(10):1876-1882. oral central venous catheter-associated deep venous thrombosis in 62. Hermann JM, Meusers M, Bachran R, et al. Self-reported regular children with diabetic ketoacidosis. Crit Care Med. 2003;31(1):80-83. alcohol consumption in adolescents and emerging adults with type 84. Worly JM, Fortenberry JD, Hansen I, Chambliss CR, Stockwell J. 1 diabetes: A neglected risk factor for diabetic ketoacidosis? Multi- Deep venous thrombosis in children with diabetic ketoacidosis and center analysis of 29 630 patients from the DPV registry. Pediatr Dia- femoral central venous catheters. Pediatrics. 2004;113(1, pt 1): betes. 2017;18(8):817-823. e57-e60. 63. Cope JU, Morrison AE, Samuels-Reid J. Adolescent use of insulin and 85. Chima RS, Hanson SJ. Venous thromboembolism in critical illness patient-controlled analgesia pump technology: a 10-year Food and and trauma: pediatric perspectives. Front Pediatr. 2017;5:47. Drug Administration retrospective study of adverse events. Pediat- 86. Flood RG, Chiang VW. Rate and prediction of infection in children rics. 2008;121(5):e1133-e1138. with diabetic ketoacidosis. Am J Emerg Med. 2001;19(4):270-273. 64. Karges B, Schwandt A, Heidtmann B, et al. Association of insulin 87. Monroe KW, King W, Atchison JA. Use of PRISM scores in triage of pump therapy vs insulin injection therapy with severe hypoglycemia, pediatric patients with diabetic ketoacidosis. Am J Manag Care. 1997; ketoacidosis, and glycemic control among children, adolescents, and 3(2):253-258. young adults with type 1 diabetes. JAMA. 2017;318(14):1358-1366. 88. Bonadio WA, Gutzeit MF, Losek JD, Smith DS. Outpatient manage- 65. Kleinman ME, Chameides L, Schexnayder SM, et al. Pediatric ment of diabetic ketoacidosis. Am J Dis Child. 1988;142(4):448-450. advanced life support: 2010 American Heart Association Guidelines 89. Linares MY, Schunk JE, Lindsay R. Laboratory presentation in dia- for Cardiopulmonary Resuscitation and Emergency Cardiovascular betic ketoacidosis and duration of therapy. Pediatr Emerg Care. 1996; Care. Pediatrics. 2010;126(5):e1361-e1399. 12(5):347-351. 66. Kleinman ME, de Caen AR, Chameides L, et al. Pediatric basic and 90. Yu HY, Agus M, Kellogg MD. Clinical utility of Abbott precision advanced life support: 2010 international consensus on cardiopulmo- Xceed pro(R) ketone meter in diabetic patients. Pediatr Diabetes. nary resuscitation and emergency cardiovascular care science with 2011;12(7):649-655. treatment recommendations. Pediatrics. 2010;126(5):e1261-e1318. 91. Lutfi R, Huang J, Wong HR. Plasmapheresis to treat hypertriglyceri- 67. Wiggam MI, O'Kane MJ, Harper R, et al. Treatment of diabetic ketoa- demia in a child with diabetic ketoacidosis and pancreatitis. Pediat- cidosis using normalization of blood 3-hydroxybutyrate concentra- rics. 2012;129(1):e195-e198. tion as the endpoint of emergency management. A randomized 92. Oh G, Anderson S, Tancredi D, Kuppermann N, Glaser N. Hyponatre- controlled study. Diabetes Care. 1997;20(9):1347-1352. mia in pediatric diabetic ketoacidosis: reevaluating the correction fac- 68. Vanelli M, Chiari G, Capuano C, Iovane B, Bernardini A, Giacalone T. tor for hyperglycemia. Arch Pediatr Adolesc Med. 2009;163(8): The direct measurement of 3-beta-hydroxy butyrate enhances the 771-772. management of diabetic ketoacidosis in children and reduces time 93. Halperin ML, Goldstein MB. Fluid, Electrolyte, and Acid-Base Physiol- and costs of treatment. Diabetes Nutr Metab. 2003;16(5–6):312-316. ogy. 3rd ed. Philadelphia, PA: Saunders; 1999. 69. Ham MR, Okada P, White PC. Bedside ketone determination in dia- 94. Love AH, Phillips RA. Measurement of dehydration in cholera. J Infect betic children with hyperglycemia and ketosis in the acute care set- Dis. 1969;119(1):39-42. ting. Pediatr Diabetes. 2004;5(1):39-43. 95. Harris GD, Fiordalisi I. Physiologic management of diabetic ketoaci- 70. Rewers A, McFann K, Chase HP. Bedside monitoring of blood demia. A 5-year prospective pediatric experience in 231 episodes. beta-hydroxybutyrate levels in the management of diabetic ketoaci- Arch Pediatr Adolesc Med. 1994;148(10):1046-1052. dosis in children. Diabetes Technol Ther. 2006;8(6):671-676. 96. Napolova O, Urbach S, Davids MR, Halperin ML. Assessing the 71. Prisco F, Picardi A, Iafusco D, et al. Blood ketone bodies in patients degree of extracellular fluid volume contraction in a patient with a with recent-onset type 1 diabetes (a multicenter study). Pediatr Dia- betes. 2006;7(4):223-228. severe degree of hyperglycaemia. Nephrol Dial Transplant. 2003; 72. Noyes KJ, Crofton P, Bath LE, et al. Hydroxybutyrate near-patient 18(12):2674-2677. testing to evaluate a new end-point for intravenous insulin therapy 97. Halperin ML, Maccari C, Kamel KS, Carlotti AP, Bohn D. Strategies to in the treatment of diabetic ketoacidosis in children. Pediatr Diabetes. diminish the danger of cerebral edema in a pediatric patient present- 2007;8(3):150-156. ing with diabetic ketoacidosis. Pediatr Diabetes. 2006;7(4):191-195. 73. Klocker AA, Phelan H, Twigg SM, Craig ME. Blood 98. Katz MA. Hyperglycemia-induced hyponatremia--calculation of beta-hydroxybutyrate vs. urine acetoacetate testing for the preven- expected serum sodium depression. N Engl J Med. 1973;289(16): tion and management of ketoacidosis in type 1 diabetes: a systematic 843-844. review. Diabet Med. 2013;30(7):818-824. 99. Hillier TA, Abbott RD, Barrett EJ. Hyponatremia: evaluating the cor- 74. Koves IH, Neutze J, Donath S, et al. The accuracy of clinical assess- rection factor for hyperglycemia. Am J Med. 1999;106(4):399-403. ment of dehydration during diabetic ketoacidosis in childhood. Diabe- 100. Harris GD, Fiordalisi I, Harris WL, Mosovich LL, Finberg L. Minimizing tes Care. 2004;27(10):2485-2487. the risk of brain herniation during treatment of diabetic ketoacide- 75. Sottosanti M, Morrison GC, Singh RN, et al. Dehydration in children mia: a retrospective and prospective study. J Pediatr. 1990;117: with diabetic ketoacidosis: a prospective study. Arch Dis Child. 2012; 22-31. 97(2):96-100. 101. Hale PM, Rezvani I, Braunstein AW, Lipman TH, Martinez N, 76. Ugale J, Mata A, Meert KL, Sarnaik AP. Measured degree of dehydra- Garibaldi L. Factors predicting cerebral edema in young children with tion in children and adolescents with type 1 diabetic ketoacidosis. diabetic ketoacidosis and new onset type I diabetes. Acta Paediatr. Pediatr Crit Care Med. 2012;13(2):e103-e107. 1997;86(6):626-631. 77. Steiner MJ, DeWalt DA, Byerley JS. Is this child dehydrated? JAMA. 102. Glaser N, Barnett P, McCaslin I, et al. Risk factors for cerebral edema 2004;291(22):2746-2754. in children with diabetic ketoacidosis. The Pediatric Emergency Med- 78. Teasdale G, Jennett B. Assessment of coma and impaired conscious- icine Collaborative Research Committee of the American Academy ness. A practical scale. Lancet. 1974;2(7872):81-84. of Pediatrics. N Engl J Med. 2001;344(4):264-269. 79. Tasker RC, Lutman D, Peters MJ. Hyperventilation in severe diabetic 103. Brown TB. Cerebral oedema in childhood diabetic ketoacidosis: is ketoacidosis. Pediatr Crit Care Med. 2005;6(4):405-411. treatment a factor? Emerg Med J. 2004;21(2):141-144. 174 WOLFSDORF ET AL.

104. Glaser NS, Ghetti S, Casper TC, Dean JM, Kuppermann N. Pediatric 126. Oh MS, Banerji MA, Carroll HJ. The mechanism of hyperchloremic diabetic ketoacidosis, fluid therapy, and cerebral injury: the design of acidosis during the recovery phase of diabetic ketoacidosis. Diabetes. a factorial randomized controlled trial. Pediatr Diabetes. 2013;14: 1981;30(4):310-313. 435-446. 127. Oh MS, Carroll HJ, Uribarri J. Mechanism of normochloremic and 105. Edge JA, Jakes RW, Roy Y, et al. The UK case-control study of cere- hyperchloremic acidosis in diabetic ketoacidosis. Nephron. 1990; bral oedema complicating diabetic ketoacidosis in children. Diabeto- 54(1):1-6. logia. 2006;49(9):2002-2009. 128. von Oettingen JE, Rhodes ET, Wolfsdorf JI. Resolution of ketoacido- 106. Kuppermann N, Ghetti S, Schunk JE, et al. Clinical trial of fluid infu- sis in children with new onset diabetes: evaluation of various defini- sion rates for pediatric diabetic ketoacidosis. N Engl J Med. 2018; tions. Diabetes Res Clin Pract. 2017;135:76-84. 378:2275-2287. 129. Chua HR, Venkatesh B, Stachowski E, et al. Plasma-Lyte 148 vs 0.9% 107. Dunger DB, Sperling MA, Acerini CL, et al. European Society for Pae- saline for fluid resuscitation in diabetic ketoacidosis. J Crit Care. diatric Endocrinology/Lawson Wilkins Pediatric Endocrine Society 2012;27(2):138-145. consensus statement on diabetic ketoacidosis in children and adoles- 130. Waldhausl W, Kleinberger G, Korn A, Dudczak R, cents. Pediatrics. 2004;113(2):e133-e140. Bratusch-Marrain P, Nowotny P. Severe hyperglycemia: effects of 108. Adrogue HJ, Barrero J, Eknoyan G. Salutary effects of modest fluid rehydration on endocrine derangements and blood glucose concen- replacement in the treatment of adults with diabetic ketoacidosis. tration. Diabetes. 1979;28(6):577-584. Use in patients without extreme volume deficit. JAMA. 1989; 131. Owen OE, Licht JH, Sapir DG. Renal function and effects of partial 262(15):2108-2113. rehydration during diabetic ketoacidosis. Diabetes. 1981;30(6): 109. Mel JM, Werther GA. Incidence and outcome of diabetic cerebral 510-518. oedema in childhood: are there predictors? J Paediatr Child Health. 132. Luzi L, Barrett EJ, Groop LC, Ferrannini E, DeFronzo RA. Metabolic 1995;31(1):17-20. effects of low-dose insulin therapy on glucose metabolism in diabetic 110. Wagner A, Risse A, Brill HL, et al. Therapy of severe diabetic ketoaci- ketoacidosis. Diabetes. 1988;37(11):1470-1477. dosis. Zero-mortality under very-low-dose insulin application. Diabe- 133. Kitabchi AE. Low-dose insulin therapy in diabetic ketoacidosis: fact tes Care. 1999;22(5):674-677. or fiction? Diabetes Metab Rev. 1989;5(4):337-363. 111. Toledo JD, Modesto V, Peinador M, et al. Sodium concentration in 134. Martin MM, Martin AA. Continuous low-dose infusion of insulin in rehydration fluids for children with ketoacidotic diabetes: effect on the treatment of diabetic ketoacidosis in children. J Pediatr. 1976; serum sodium concentration. J Pediatr. 2009;154:895-900. 89(4):560-564. 112. White PC, Dickson BA. Low morbidity and mortality in children with 135. Edwards GA, Kohaut EC, Wehring B, Hill LL. Effectiveness of diabetic ketoacidosis treated with isotonic fluids. J Pediatr. 2013; low-dose continuous intravenous insulin infusion in diabetic ketoaci- 163(3):761-766. dosis. A prospective comparative study. J Pediatr. 1977;91(5): 113. Nallasamy K, Jayashree M, Singhi S, Bansal A. Low-dose vs 701-705. standard-dose insulin in pediatric diabetic ketoacidosis: a randomized 136. Drop SL, Duval-Arnould JM, Gober AE, Hersh JH, McEnery PT, clinical trial. JAMA Pediatr. 2014;168:999-1005. Knowles HC. Low-dose intravenous insulin infusion versus subcuta- 114. Yung M, Letton G, Keeley S. Controlled trial of Hartmann's solution neous insulin injection: a controlled comparative study of diabetic versus 0.9% saline for diabetic ketoacidosis. J Paediatr Child Health. ketoacidosis. Pediatrics. 1977;59(5):733-738. 2017;53(1):12-17. 137. Lightner ES, Kappy MS, Revsin B. Low-dose intravenous insulin infu- 115. Rother KI, Schwenk WF 2nd. Effect of rehydration fluid with sion in patients with diabetic ketoacidosis: biochemical effects in 75 mmol/L of sodium on serum sodium concentration and serum children. Pediatrics. 1977;60(5):681-688. osmolality in young patients with diabetic ketoacidosis. Mayo Clin 138. Perkin RM, Marks JF. Low-dose continuous intravenous insulin infu- Proc. 1994;69(12):1149-1153. sion in childhood diabetic ketoacidosis. Clin Pediatr (Phila). 1979; 116. Felner EI, White PC. Improving management of diabetic ketoacidosis 18(9):540-545-8. in children. Pediatrics. 2001;108(3):735-740. 139. Kappy MS, Lightner ES. Low-dose intravenous insulin in the treat- 117. Basnet S, Venepalli PK, Andoh J, Verhulst S, Koirala J. Effect of nor- ment of diabetic ketoacidosis. Am J Dis Child. 1979;133(5):523-525. mal saline and half normal saline on serum electrolytes during recov- 140. Burghen GA, Etteldorf JN, Fisher JN, Kitabchi AQ. Comparison of ery phase of diabetic ketoacidosis. J Intensive Care Med. 2014;29(1): high-dose and low-dose insulin by continuous intravenous infusion 38-42. 118. Fiordalisi I, Novotny WE, Holbert D, Finberg L, Harris GD. An 18-yr in the treatment of diabetic ketoacidosis in children. Diabetes Care. prospective study of pediatric diabetic ketoacidosis: an approach to 1980;3(1):15-20. minimizing the risk of brain herniation during treatment. Pediatr Dia- 141. Kitabchi AE, Umpierrez GE, Fisher JN, Murphy MB, Stentz FB. Thirty betes. 2007;8(3):142-149. years of personal experience in hyperglycemic crises: diabetic ketoa- 119. Hoorn EJ, Carlotti AP, Costa LA, et al. Preventing a drop in effective cidosis and hyperglycemic hyperosmolar state. J Clin Endocrinol plasma osmolality to minimize the likelihood of cerebral edema dur- Metab. 2008;93(5):1541-1552. ing treatment of children with diabetic ketoacidosis. J Pediatr. 2007; 142. Lindsay R, Bolte RG. The use of an insulin bolus in low-dose insulin 150(5):467-473. infusion for pediatric diabetic ketoacidosis. Pediatr Emerg Care. 1989; 120. Duck SC, Wyatt DT. Factors associated with brain herniation in the 5(2):77-79. treatment of diabetic ketoacidosis. J Pediatr. 1988;113:10-14. 143. Van der Meulen JA, Klip A, Grinstein S. Possible mechanism for cere- 121. Adrogue HJ, Wilson H, Boyd AE 3rd, Suki WN, Eknoyan G. Plasma bral oedema in diabetic ketoacidosis. Lancet. 1987;2(8554):306-308. acid-base patterns in diabetic ketoacidosis. N Engl J Med. 1982; 144. Soler NG, FitzGerald MG, Wright AD, Malins JM. Comparative study 307(26):1603-1610. of different insulin regimens in management of diabetic ketoacidosis. 122. Adrogue HJ, Eknoyan G, Suki WK. Diabetic ketoacidosis: role of the Lancet. 1975;2(7947):1221-1224. kidney in the acid-base homeostasis re-evaluated. Kidney Int. 1984; 145. Puttha R, Cooke D, Subbarayan A, et al. Low dose (0.05 units/kg/h) 25(4):591-598. is comparable with standard dose (0.1 units/kg/h) intravenous insulin 123. Taylor D, Durward A, Tibby SM, et al. The influence of hyperchlorae- infusion for the initial treatment of diabetic ketoacidosis in children mia on acid base interpretation in diabetic ketoacidosis. Intensive with type 1 diabetes-an observational study. Pediatr Diabetes. 2010; Care Med. 2006;32(2):295-301. 11:12-17. 124. Durward A, Skellett S, Mayer A, Taylor D, Tibby SM, Murdoch IA. 146. Al Hanshi S, Shann F. Insulin infused at 0.05 versus 0.1 units/kg/hr The value of the chloride: sodium ratio in differentiating the aetiol- in children admitted to intensive care with diabetic ketoacidosis. ogy of metabolic acidosis. Intensive Care Med. 2001;27(5):828-835. Pediatr Crit Care Med. 2011;12(2):137-140. 125. Oh MS, Carroll HJ, Goldstein DA, Fein IA. Hyperchloremic acidosis 147. Wang J, Barbry P, Maiyar AC, et al. SGK integrates insulin and miner- during the recovery phase of diabetic ketosis. Ann Intern Med. 1978; alocorticoid regulation of epithelial sodium transport. Am J Physiol 89(6):925-927. Renal Physiol. 2001;280(2):F303-F313. WOLFSDORF ET AL. 175

148. Tallini NY, Stoner LC. Amiloride-sensitive sodium current in everted 171. Keller U, Berger W. Prevention of hypophosphatemia by phosphate Ambystoma initial collecting tubule: short-term insulin effects. infusion during treatment of diabetic ketoacidosis and hyperosmolar Am J Physiol Cell Physiol. 2002;283(4):C1171-C1181. coma. Diabetes. 1980;29(2):87-95. 149. McCormick JA, Bhalla V, Pao AC, Pearce D. SGK1: a rapid 172. Wilson HK, Keuer SP, Lea AS, Boyd AE 3rd, Eknoyan G. Phosphate aldosterone-induced regulator of renal sodium reabsorption. Phys therapy in diabetic ketoacidosis. Arch Intern Med. 1982;142(3): Ther. 2005;20:134-139. 517-520. 150. Tiwari S, Nordquist L, Halagappa VK, Ecelbarger CA. Trafficking of 173. Becker DJ, Brown DR, Steranka BH, Drash AL. Phosphate replace- ENaC subunits in response to acute insulin in mouse kidney. ment during treatment of diabetic ketosis. Effects on calcium and Am J Physiol Renal Physiol. 2007;293(1):F178-F185. phosphorus homeostasis. Am J Dis Child. 1983;137(3):241-246. 151. Frindt G, Palmer LG. Effects of insulin on Na and K transporters in 174. Fisher JN, Kitabchi AE. A randomized study of phosphate therapy in the rat CCD. Am J Physiol Renal Physiol. 2012;302(10):F1227-F1233. the treatment of diabetic ketoacidosis. J Clin Endocrinol Metab. 1983; 152. Carlotti AP, St George-Hyslop C, Bohn D, Halperin ML. Hypokalemia 57(1):177-180. during treatment of diabetic ketoacidosis: clinical evidence for an 175. Clerbaux T, Reynaert M, Willems E, Frans A. Effect of phosphate on aldosterone-like action of insulin. J Pediatr. 2013;163(1):207-212.e1. oxygen-hemoglobin affinity, diphosphoglycerate and blood gases 153. Fisher JN, Shahshahani MN, Kitabchi AE. Diabetic ketoacidosis: during recovery from diabetic ketoacidosis. Intensive Care Med. low-dose insulin therapy by various routes. N Engl J Med. 1977; 1989;15(8):495-498. 297(5):238-241. 176. Weisinger JR, Bellorin-Font E. Magnesium and phosphorus. Lancet. 154. Sacks HS, Shahshahani M, Kitabchi AE, Fisher JN, Young RT. Similar 1998;352(9125):391-396. responsiveness of diabetic ketoacidosis to low-dose insulin by intra- 177. Singhal PC, Kumar A, Desroches L, Gibbons N, Mattana J. Prevalence muscular injection and albumin-free infusion. Ann Intern Med. 1979; and predictors of rhabdomyolysis in patients with hypophosphate- 90(1):36-42. mia. Am J Med. 1992;92(5):458-464. 155. Umpierrez GE, Latif K, Stoever J, et al. Efficacy of subcutaneous 178. Kutlu AO, Kara C, Cetinkaya S. Rhabdomyolysis without detectable insulin lispro versus continuous intravenous regular insulin for the myoglobulinuria due to severe hypophosphatemia in diabetic ketoa- treatment of patients with diabetic ketoacidosis. Am J Med. 2004; cidosis. Pediatr Emerg Care. 2011;27(6):537-538. 179. Bohannon NJ. Large phosphate shifts with treatment for hyperglyce- 117(5):291-296. 156. Umpierrez GE, Cuervo R, Karabell A, Latif K, Freire AX, Kitabchi AE. mia. Arch Intern Med. 1989;149(6):1423-1425. 180. de Oliveira Iglesias SB, Pons Leite H, de Carvalho WB. Treatment of diabetic ketoacidosis with subcutaneous insulin aspart. Hypophosphatemia-induced seizure in a child with diabetic ketoaci- Diabetes Care. 2004;27(8):1873-1878. dosis. Pediatr Emerg Care. 2009;25(12):859-861. 157. Della Manna T, Steinmetz L, Campos PR, et al. Subcutaneous use of 181. Zipf WB, Bacon GE, Spencer ML, Kelch RP, Hopwood NJ, a fast-acting insulin analog: an alternative treatment for pediatric Hawker CD. Hypocalcemia, hypomagnesemia, and transient hypo- patients with diabetic ketoacidosis. Diabetes Care. 2005;28(8): parathyroidism during therapy with potassium phosphate in diabetic 1856-1861. ketoacidosis. Diabetes Care. 1979;2(3):265-268. 158. Savoldelli RD, Farhat SC, Manna TD. Alternative management of dia- 182. Winter RJ, Harris CJ, Phillips LS, Green OC. Diabetic ketoacidosis. betic ketoacidosis in a Brazilian pediatric emergency department. Induction of hypocalcemia and hypomagnesemia by phosphate ther- Diabetol Metab Syndr. 2010;2:41. apy. Am J Med. 1979;67(5):897-900. 159. Cohen M, Leibovitz N, Shilo S, Zuckerman-Levin N, Shavit I, 183. Hale PJ, Crase J, Nattrass M. Metabolic effects of bicarbonate in the Shehadeh N. Subcutaneous regular insulin for the treatment of dia- treatment of diabetic ketoacidosis. Br Med J (Clin Res Ed). 1984; betic ketoacidosis in children. Pediatr Diabetes. 2017;18(4):290-296. 289(6451):1035-1038. 160. Adrogue HJ, Lederer ED, Suki WN, Eknoyan G. Determinants of 184. Morris LR, Murphy MB, Kitabchi AE. Bicarbonate therapy in severe plasma potassium levels in diabetic ketoacidosis. Medicine diabetic ketoacidosis. Ann Intern Med. 1986;105(6):836-840. (Baltimore). 1986;65(3):163-172. 185. Okuda Y, Adrogue HJ, Field JB, Nohara H, Yamashita K. Counterpro- 161. DeFronzo RA, Felig P, Ferrannini E, Wahren J. Effect of graded doses ductive effects of sodium bicarbonate in diabetic ketoacidosis. J Clin of insulin on splanchnic and peripheral potassium metabolism in Endocrinol Metab. 1996;81(1):314-320. man. Am J Phys. 1980;238(5):E421-E427. 186. Green SM, Rothrock SG, Ho JD, et al. Failure of adjunctive bicarbon- 162. Tattersall RB. A paper which changed clinical practice (slowly). Jacob ate to improve outcome in severe pediatric diabetic ketoacidosis. holler on potassium deficiency in diabetic acidosis (1946). Diabet Ann Emerg Med. 1998;31(1):41-48. Med. 1999;16(12):978-984. 187. Assal JP, Aoki TT, Manzano FM, Kozak GP. Metabolic effects of 163. Davis SM, Maddux AB, Alonso GT, Okada CR, Mourani PM, sodium bicarbonate in management of diabetic ketoacidosis. Diabe- Maahs DM. Profound hypokalemia associated with severe diabetic tes. 1974;23(5):405-411. ketoacidosis. Pediatr Diabetes. 2016;17:61-65. 188. Ohman JL Jr, Marliss EB, Aoki TT, Munichoodappa CS, Khanna VV, 164. Guest G. Organic phosphates of the blood and mineral metabolism in Kozak GP. The cerebrospinal fluid in diabetic ketoacidosis. N Engl J diabetic acidosis. Am J Dis Child. 1942;64:401-412. Med. 1971;284(6):283-290. 165. Guest G, Rapoport S. Electrolytes of blood plasma and cells in dia- 189. Soler NG, Bennett MA, Dixon K, FitzGerald MG, Malins JM. Potas- betic acidosis and during recovery. Proc Amer Diab Assoc. 1947;7: sium balance during treatment of diabetic ketoacidosis with special 95-115. reference to the use of bicarbonate. Lancet. 1972;2(7779):665-667. 166. Riley MS, Schade DS, Eaton RP. Effects of insulin infusion on plasma 190. Lever E, Jaspan JB. Sodium bicarbonate therapy in severe diabetic phosphate in diabetic patients. Metabolism. 1979;28(3):191-194. ketoacidosis. Am J Med. 1983;75(2):263-268. 167. Alberti KG, Emerson PM, Darley JH, Hockaday TD. 191. Narins RG, Cohen JJ. Bicarbonate therapy for organic acidosis: the 2,3-Diphosphoglycerate and tissue oxygenation in uncontrolled dia- case for its continued use. Ann Intern Med. 1987;106(4):615-618. betes mellitus. Lancet. 1972;2(7774):391-395. 192. Shankar V, Haque A, Churchwell KB, Russell W. Insulin glargine sup- 168. Knochel JP. The pathophysiology and clinical characteristics of plementation during early management phase of diabetic ketoacido- severe hypophosphatemia. Arch Intern Med. 1977;137(2):203-220. sis in children. Intensive Care Med. 2007;33(7):1173-1178. 169. O'Connor LR, Wheeler WS, Bethune JE. Effect of hypophosphatemia 193. Harrison VS, Rustico S, Palladino AA, Ferrara C, Hawkes CP. Glargine on myocardial performance in man. N Engl J Med. 1977;297(17): co-administration with intravenous insulin in pediatric diabetic 901-903. ketoacidosis is safe and facilitates transition to a subcutaneous regi- 170. Gibby OM, Veale KE, Hayes TM, Jones JG, Wardrop CA. Oxygen men. Pediatr Diabetes. 2017;18(8):742-748. availability from the blood and the effect of phosphate replacement 194. Curtis JR, To T, Muirhead S, Cummings E, Daneman D. Recent trends on erythrocyte 2,3-diphosphoglycerate and haemoglobin-oxygen in hospitalization for diabetic ketoacidosis in Ontario children. Diabe- affinity in diabetic ketoacidosis. Diabetologia. 1978;15(5):381-385. tes Care. 2002;25(9):1591-1596. 176 WOLFSDORF ET AL.

195. Edge JA, Ford-Adams ME, Dunger DB. Causes of death in children 217. Glaser NS, Wootton-Gorges SL, Buonocore MH, et al. Frequency of with insulin dependent diabetes 1990-96. Arch Dis Child. 1999;81(4): sub-clinical cerebral edema in children with diabetic ketoacidosis. 318-323. Pediatr Diabetes. 2006;7(2):75-80. 196. Decourcey DD, Steil GM, Wypij D, Agus MS. Increasing use of 218. Glaser NS, Marcin JP, Wootton-Gorges SL, et al. Correlation of clini- hypertonic saline over mannitol in the treatment of symptomatic cal and biochemical findings with diabetic ketoacidosis-related cere- cerebral edema in pediatric diabetic ketoacidosis: an 11-year retro- bral edema in children using magnetic resonance diffusion-weighted spective analysis of mortality. Pediatr Crit Care Med. 2013;14(7): imaging. J Pediatr. 2008;153(4):541-546. 694-700. 219. Krane EJ, Rockoff MA, Wallman JK, Wolfsdorf JI. Subclinical brain 197. Saydah SH, Imperatore G, Geiss LS, Gregg E. Diabetes death rates swelling in children during treatment of diabetic ketoacidosis. N Engl among youths aged ≤19 years: United States, 1968–2009. Morb J Med. 1985;312(18):1147-1151. Mortal Wkly Rep. 2012;61(43):869-872. 220. Hoffman WH, Steinhart CM, el Gammal T, Steele S, Cuadrado AR, 198. Morgan E, Black CR, Abid N, Cardwell CR, McCance DR, Morse PK. Cranial CT in children and adolescents with diabetic ketoacidosis. AJNR Am J Neuroradiol. 1988;9(4):733-739. Patterson CC. Mortality in type 1 diabetes diagnosed in childhood in 221. Sperling MA. Cerebral edema in diabetic ketoacidosis: an underesti- Northern Ireland during 1989-2012: A population-based cohort mated complication? Pediatr Diabetes. 2006;7(2):73-74. study. Pediatr Diabetes. 2018;19(1):166-170. 222. Arieff AI, Kleeman CR. Studies on mechanisms of cerebral edema in 199. Gibb FW, Teoh WL, Graham J, Lockman KA. Risk of death following diabetic comas. Effects of hyperglycemia and rapid lowering of admission to a UK hospital with diabetic ketoacidosis. Diabetologia. plasma glucose in normal rabbits. J Clin Invest. 1973;52(3):571-583. 2016;59(10):2082-2087. 223. Arieff AI, Kleeman CR. Cerebral edema in diabetic comas. II. Effects 200. Sperling MA. Diabetes: recurrent DKA – for whom the bell tolls. Nat of hyperosmolality, hyperglycemia and insulin in diabetic rabbits. J Rev Endocrinol. 2016;12(10):562-564. Clin Endocrinol Metab. 1974;38(6):1057-1067. 201. Daneman D. Diabetes-related mortality. A pediatrician's view. Diabe- 224. Prockop LD. Hyperglycemia, polyol accumulation, and increased tes Care. 2001;24(5):801-802. intracranial pressure. Arch Neurol. 1971;25(2):126-140. 202. Edge JA, Hawkins MM, Winter DL, Dunger DB. The risk and out- 225. Harris GD, Fiordalisi I, Finberg L. Safe management of diabetic ketoa- come of cerebral oedema developing during diabetic ketoacidosis. cidemia. J Pediatr. 1988;113:65-67. Arch Dis Child. 2001;85(1):16-22. 226. Kamel KS, Halperin ML. Acid-base problems in diabetic ketoacidosis. 203. Lawrence SE, Cummings EA, Gaboury I, Daneman D. N Engl J Med. 2015;372(6):546-554. Population-based study of incidence and risk factors for cerebral 227. Glaser NS, Wootton-Gorges SL, Marcin JP, et al. Mechanism of cere- edema in pediatric diabetic ketoacidosis. J Pediatr. 2005;146(5): bral edema in children with diabetic ketoacidosis. J Pediatr. 2004; 688-692. 145(2):164-171. 204. Ghetti S, Lee JK, Sims CE, Demaster DM, Glaser NS. Diabetic ketoa- 228. Glaser N, Yuen N, Anderson SE, Tancredi DJ, O'Donnell ME. Cere- cidosis and memory dysfunction in children with type 1 diabetes. J bral metabolic alterations in rats with diabetic ketoacidosis: effects Pediatr. 2010;156(1):109-114. of treatment with insulin and intravenous fluids and effects of bume- 205. Nadebaum C, Scratch SE, Northam EA, Cameron FJ, Diabetic Ketoa- tanide. Diabetes. 2010;59(3):702-709. cidosis and Brain Injury Study Group. Clinical utility of mental state 229. Yuen N, Anderson SE, Glaser N, Tancredi DJ, O'Donnell ME. Cere- screening as a predictor of intellectual outcomes 6 months after bral blood flow and cerebral edema in rats with diabetic ketoacidosis. diagnosis of type 1 diabetes. Pediatr Diabetes. 2012;13(8):632-637. Diabetes. 2008;57(10):2588-2594. 206. Cameron FJ, Scratch SE, Nadebaum C, et al. Neurological conse- 230. Tasker RC, Acerini CL. Cerebral edema in children with diabetic quences of diabetic ketoacidosis at initial presentation of type 1 dia- ketoacidosis: vasogenic rather than cellular? Pediatr Diabetes. 2014; betes in a prospective cohort study of children. Diabetes Care. 2014; 15(4):261-270. 37(6):1554-1562. 231. Rosenbloom AL, Schatz DA, Krischer JP, et al. Therapeutic contro- 207. Shalitin S, Fisher S, Yackbovitch-Gavan M, et al. Ketoacidosis at versy: prevention and treatment of diabetes in children. J Clin Endo- onset of type 1 diabetes is a predictor of long-term glycemic control. crinol Metab. 2000;85(2):494-522. Pediatr Diabetes. 2018;19(2):320-328. 232. Glaser N. Cerebral injury and cerebral edema in children with dia- 208. Duca LM, Wang B, Rewers M, Rewers A. Diabetic ketoacidosis at betic ketoacidosis: could cerebral ischemia and reperfusion injury be diagnosis of type 1 diabetes predicts poor long-term glycemic con- involved? Pediatr Diabetes. 2009;10(8):534-541. trol. Diabetes Care. 2017;40(9):1249-1255. 233. Hoffman WH, Stamatovic SM, Andjelkovic AV. Inflammatory media- 209. Yasuda K, Hayashi M, Murayama M, Yamakita N. Acidosis-induced tors and blood brain barrier disruption in fatal brain edema of dia- betic ketoacidosis. Brain Res. 2009;1254:138-148. hypochloremic alkalosis in diabetic ketoacidosis confirmed by the 234. Vavilala MS, Richards TL, Roberts JS, et al. Change in blood-brain modified base excess method. J Clin Endocrinol Metab. 2016;101(6): barrier permeability during pediatric diabetic ketoacidosis treatment. 2390-2395. Pediatr Crit Care Med. 2010;11(3):332-338. 210. Cooper MR, Turner RA Jr, Hutaff L, Prichard R. Diabetic 235. Rosenbloom AL. Intracerebral crises during treatment of diabetic keto-acidosis complicated by disseminated intravascular coagulation. ketoacidosis. Diabetes Care. 1990;13(1):22-33. South Med J. 1973;66(6):653-657. 236. Bello FA, Sotos JF. Cerebral oedema in diabetic ketoacidosis in chil- 211. Keane S, Gallagher A, Ackroyd S, McShane MA, Edge JA. Cerebral dren [letter]. Lancet. 1990;336(8706):64. venous thrombosis during diabetic ketoacidosis. Arch Dis Child. 237. Mahoney CP, Vlcek BW, DelAguila M. Risk factors for developing 2002;86(3):204-205. brain herniation during diabetic ketoacidosis. Pediatr Neurol. 1999; 212. Ho J, Mah JK, Hill MD, Pacaud D. Pediatric stroke associated with 21(4):721-727. new onset type 1 diabetes mellitus: case reports and review of the 238. Durr JA, Hoffman WH, Sklar AH, el Gammal T, Steinhart CM. Corre- literature. Pediatr Diabetes. 2006;7(2):116-121. lates of brain edema in uncontrolled IDDM. Diabetes. 1992;41(5): 213. Moye J, Rosenbloom AL, Silverstein J. Clinical predictors of mucor- 627-632. mycosis in children with type 1 diabetes mellitus. J Pediatr Endocrinol 239. Durward A, Ferguson LP, Taylor D, Murdoch IA, Tibby SM. The tem- Metab. 2002;15(7):1001-1004. poral relationship between glucose-corrected serum sodium and 214. Watson JP, Barnett AH. Pneumomediastinum in diabetic ketoacido- neurological status in severe diabetic ketoacidosis. Arch Dis Child. sis. Diabet Med. 1989;6(2):173-174. 2011;96:50-57. 215. Hursh BE, Ronsley R, Islam N, Mammen C, Panagiotopoulos C. Acute 240. Bureau MA, Begin R, Berthiaume Y, Shapcott D, Khoury K, kidney injury in children with type 1 diabetes hospitalized for dia- Gagnon N. Cerebral hypoxia from bicarbonate infusion in diabetic betic ketoacidosis. JAMA Pediatr. 2017;171(5):e170020. acidosis. J Pediatr. 1980;96(6):968-973. 216. Haddad NG, Croffie JM, Eugster EA. Pancreatic enzyme elevations in 241. Deeb L. Development of fatal cerebral edema during outpatient ther- children with diabetic ketoacidosis. J Pediatr. 2004;145(1):122-124. apy for diabetic ketoacidosis. Pract Diab. 1989;6:212-213. WOLFSDORF ET AL. 177

242. Glasgow AM. Devastating cerebral edema in diabetic ketoacidosis 259. Matz R. Management of the hyperosmolar hyperglycemic syndrome. before therapy [letter]. Diabetes Care. 1991;14(1):77-78. Am Fam Physician. 1999;60(5):1468-1476. 243. Couch RM, Acott PD, Wong GW. Early onset fatal cerebral edema in 260. Delaney MF, Zisman A, Kettyle WM. Diabetic ketoacidosis and diabetic ketoacidosis. Diabetes Care. 1991;14(1):78-79. hyperglycemic hyperosmolar nonketotic syndrome. Endocrinol Metab 244. Fiordalisi I, Harris GD, Gilliland MG. Prehospital cardiac arrest in dia- Clin N Am. 2000;29(4):683-705. betic ketoacidemia: why brain swelling may lead to death before 261. Mannix R, Tan ML, Wright R, Baskin M. Acute pediatric rhabdomyolysis: treatment. J Diabetes Complicat. 2002;16(3):214-219. causes and rates of renal failure. Pediatrics. 2006;118(5):2119-2125. 245. Edge JA. Cerebral oedema during treatment of diabetic ketoacidosis: 262. Hollander AS, Olney RC, Blackett PR, Marshall BA. Fatal malignant are we any nearer finding a cause? Diabetes Metab Res Rev. 2000; hyperthermia-like syndrome with rhabdomyolysis complicating the 16(5):316-324. presentation of diabetes mellitus in adolescent males. Pediatrics. 246. Muir AB, Quisling RG, Yang MC, Rosenbloom AL. Cerebral edema in 2003;111(6, pt 1):1447-1452. childhood diabetic ketoacidosis: natural history, radiographic find- 263. Carcillo JA, Tasker RC. Fluid resuscitation of hypovolemic shock: ings, and early identification. Diabetes Care. 2004;27(7):1541-1546. acute medicine's great triumph for children. Intensive Care Med. 247. Franklin B, Liu J, Ginsberg-Fellner F. Cerebral edema and ophthalmo- 2006;32(7):958-961. plegia reversed by mannitol in a new case of insulin-dependent dia- 264. Kilbane BJ, Mehta S, Backeljauw PF, Shanley TP, Crimmins NA. betes mellitus. Pediatrics. 1982;69(1):87-90. Approach to management of malignant hyperthermia-like syndrome in 248. Shabbir N, Oberfield SE, Corrales R, Kairam R, Levine LS. Recovery pediatric diabetes mellitus. Pediatr Crit Care Med. 2006;7(2):169-173. from symptomatic brain swelling in diabetic ketoacidosis. Clin Pediatr 265. Laffel LM, Wentzell K, Loughlin C, Tovar A, Moltz K, Brink S. Sick day (Phila). 1992;31(9):570-573. management using blood 3-hydroxybutyrate (3-OHB) compared with 249. Roberts MD, Slover RH, Chase HP. Diabetic ketoacidosis with intra- urine ketone monitoring reduces hospital visits in young people with cerebral complications. Pediatr Diabetes. 2001;2:109-114. T1DM: a randomized clinical trial. Diabet Med. 2006;23(3):278-284. 250. Curtis JR, Bohn D, Daneman D. Use of hypertonic saline in the treat- 266. Laffel L. Sick-day management in type 1 diabetes. Endocrinol Metab ment of cerebral edema in diabetic ketoacidosis (DKA). Pediatr Diabe- Clin N Am. 2000;29(4):707-723. tes. 2001;2(4):191-194. 267. Hoffman WH, O'Neill P, Khoury C, Bernstein SS. Service and education 251. Kamat P, Vats A, Gross M, Checchia PA. Use of hypertonic saline for for the insulin-dependent child. Diabetes Care. 1978;1(5):285-288. the treatment of altered mental status associated with diabetic 268. Drozda DJ, Dawson VA, Long DJ, Freson LS, Sperling MA. Assess- ketoacidosis. Pediatr Crit Care Med. 2003;4(2):239-242. ment of the effect of a comprehensive diabetes management pro- 252. Tasker RC, Burns J. Hypertonic saline therapy for cerebral edema in gram on hospital admission rates of children with diabetes mellitus. diabetic ketoacidosis: no change yet, please. Pediatr Crit Care Med. Diabetes Educ. 1990;16(5):389-393. 2014;15(3):284-285. 269. Grey M, Boland EA, Davidson M, Li J, Tamborlane WV. Coping 253. Soto-Rivera CL, Asaro LA, Agus MS, DeCourcey DD. Suspected skills training for youth with diabetes mellitus has long-lasting effects cerebral edema in diabetic ketoacidosis: is there still a role for head on metabolic control and quality of life. JPediatr. 2000;137(1):107-113. CT in treatment decisions? Pediatr Crit Care Med. 2017;18(3): 270. Golden MP, Herrold AJ, Orr DP. An approach to prevention of recur- 207-212. rent diabetic ketoacidosis in the pediatric population. J Pediatr. 1985; 254. Kanter RK, Oliphant M, Zimmerman JJ, Stuart MJ. Arterial thrombo- 107(2):195-200. sis causing cerebral edema in association with diabetic ketoacidosis. 271. Pinhas-Hamiel O, Sperling M. Diabetic ketoacidosis. In: Hochberg Z, Crit Care Med. 1987;15(2):175-176. ed. Practical Algorithms in Pediatric Endocrinology. 3rd ed. Basel, Swit- 255. Roe TF, Crawford TO, Huff KR, Costin G, Kaufman FR, Nelson MD zerland: Karger; 2017:112-113. Jr. Brain infarction in children with diabetic ketoacidosis. J Diabetes Complicat. 1996;10(2):100-108. 256. Rosenbloom AL. Fatal cerebral infarctions in diabetic ketoacidosis in a child with previously unknown heterozygosity for factor V Leiden How to cite this article: Wolfsdorf JI, Glaser N, Agus M, et al. deficiency. J Pediatr. 2004;145(4):561-562. 257. Fourtner SH, Weinzimer SA, Levitt Katz LE. Hyperglycemic hyperos- ISPAD Clinical Practice Consensus Guidelines 2018: Diabetic molar non-ketotic syndrome in children with type 2 diabetes. Pediatr ketoacidosis and the hyperglycemic hyperosmolar state. Diabetes. 2005;6(3):129-135. Pediatr Diabetes. 2018;19(Suppl. 27):155–177. https://doi. 258. Kronan K, Normal ME. Renal and electrolyte emergencies. In: org/10.1111/pedi.12701 Fleisher GR, Ludwig S, eds. Textbook of Emergency Medicine. 4th ed. Philadelphia, PA: Lippincott, Williams and Wilkins; 2000.